Editor-in-Chief

Dr. Jose Navarro Pedreño

University Miguel Hernández of Elche, Spain

Associate Editor

Prof. Kaiyong Wang

Chinese Academy of Sciences, China

Editorial Board Members

Aleksandar Djordje Valjarević, Serbia Muhammad Asif, Pakistan

Sanwei He, China Nevin Özdemir, Turkey
Christos Kastrisios, United Marwan Ghaleb Ghanem, Palestinian
Fei Li, China Liqiang Zhang, China
Adeline NGIE, South Africa Bodo Tombari, Nigeria
Zhixiang Fang, China Lingyue LI, China
Rubén Camilo Lois-González, Spain John P. Tiefenbacher, United States
Jesús López-Rodríguez, Spain Mehmet Cetin, Turkey
Keith Hollinshead, United Kingdom Najat Qader Omar, IRAQ
Rudi Hartmann, United States Binod Dawadi, Nepal
Mirko Andreja Borisov, Serbia Julius Oluranti Owoeye, Nigeria
Ali Hosseini, Iran Carlos Teixeira, Canada
Kaiyong Wang, China James Kurt Lein, Greece
Virginia Alarcón Martínez, Spain Angel Paniagua Mazorra, Spain
Krystle Ontong, South Africa Ola Johansson, United States
Jesús M. González-Pérez, Spain John Manyimadin Kusimi, Ghana
Pedro Robledo Ardila, Spain Susan Ihuoma Ajiere, Nigeria
Federico R. León, Peru Zhiguo Yao,China
Eva Savina Malinverni, Italy Shengpei Dai, China
Alexander Standish, United Kingdom Diego Giuliani, Italy
Samson Olaitan Olanrewaju, Nigeria Lede Niu, China
Zhibao Wang, China Zhen Liu, China
Levent Yilmaz, Turkey Chengpeng Lu, China
Kecun Zhang,China Haoming Xia, China
Cheikh Faye, Senegal Yanbin Chen, China
Chiara Certomà, Italy Zhonglei Yu, China
Christopher Robin Bryant, Canada
 Volume 6 Issue 3 • July 2023 • ISSN 2630-5070 (Online)

Journal of
Geographical Research
Editor-in-Chief
Dr. Jose Navarro Pedreño
 Volume 6 ｜ Issue 3 ｜ July 2023 ｜ Page1-80
Journal of Geographical Research

Contents
Articles
1 Combating Climate Change and Global Warming for a Sustainable Living in Harmony with Nature
V. Balaram
18 Spatiotemporal Analysis of Land Use Land Cover Mapping and Change Detection in Dambatta Local
Government Area
David Sesugh Aule, Mamman Saba Jibril, Ali Hussain Idris
29 Application of Vegetation Indices for Detection and Monitoring Oil Spills in Ahoada West Local Govern-
ment Area of Rivers State, Nigeria
Jonathan Lisa Erebi, Egirani E. Davidson
42 Considering Regional Connectivity and Policy Factors in the Simulation of Land Use Change in New
Areas: A Case Study of Nansha New District, China
Zehuan Zheng, Shi Xian
61 Spatial Agglomeration and Diffusion of Population Based on a Regional Density Function Approach:
A Case Study of Shandong Province in China
Xiaohan Zhao, Yanbin Chen
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Journal of Geographical Research

https://journals.bilpubgroup.com/index.php/jgr

ARTICLE

Combating Climate Change and Global Warming for a Sustainable

Living in Harmony with Nature
V. Balaram

CSIR—National Geophysical Research Institute, Hyderabad, 500007, India

ABSTRACT
As humanity has been polluting the atmosphere with greenhouse gases, the planet is getting warmed up which
is triggering the frequency and the intensity of extreme events like heat waves, dry conditions, wildfires, cyclones,
tornadoes, lightning, and massive flooding all over the planet Earth. There is considerable evidence that the
concentration of greenhouse gases, especially that of CO2 has steadily increased in the atmosphere as a result of
the indiscriminate use of fossil fuels around the world particularly during the last 70 years. The glaciers in the high
mountain and polar regions are diminishing fast, sea levels are rising, and food production is being affected severely
in certain parts of the world. In fact, the changing climate has currently become one of the major threats to the survival
of civilization. The world scientific communities are warning of a climate emergency and requesting the decision
makers to promptly respond and act to sustain life on planet Earth. To deliver net zero emissions by the year 2050, the
whole world must phase out the technologies such as coal-powered thermal plants and diesel/petrol/gasoline-powered
vehicles which release abundant amounts of CO2 and other greenhouse gases into the atmosphere and invest in the
development of clean energies such as hydel, wind, solar, space-solar, and nuclear energies. This transition to a low-
carbon economy with the help of these technologies together with other technologies such as hydrogen fuel, fuel cells,
electric vehicles, and massive plantations is expected to take our planet Earth to a safe zone in the coming 20-30 years.
Keywords: Greenhouse gases; Ocean acidification; Diminishing of glaciers; Sea level rise; Carbon capture; Direct air
capture; Critical minerals in deep-sea

*CORRESPONDING AUTHOR:
Balaram, V., CSIR—National Geophysical Research Institute, Hyderabad, 500007, India; Email: [email protected]
ARTICLE INFO
Received: 17 April 2023 | Revised: 19 June 2023 | Accepted: 25 June 2023 | Published Online: 29 June 2023
DOI: https://doi.org/10.30564/jgr.v6i3.5706
CITATION
Balaram, V., 2023. Combating Climate Change and Global Warming for a Sustainable Living in Harmony with Nature. Journal of Geographical
Research. 6(3): 1-17. DOI: https://doi.org/10.30564/jgr.v6i3.5706
COPYRIGHT
Copyright © 2023 by the author(s). Published by Bilingual Publishing Group. This is an open access article under the Creative Commons Attribu-
tion-NonCommercial 4.0 International (CC BY-NC 4.0) License. (https://creativecommons.org/licenses/by-nc/4.0/).

1
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

1. Introduction tures of the surface of the earth as well as oceans and

their projected continuation due to the greenhouse
Our planet is very complex and delicate in sev- effect. Further deeper studies indicated that the
eral ways, without taking proper steps to preserve
other general processes such as variations in solar
its ecosystem related to life on Earth would not
radiation, deviations in the Earth’s orbit, volcanic
sustain itself. Currently, global warming is one of
eruptions, mountain-building and continental drift,
the biggest threats to the survival of the natural
and changes in GHG concentrations are responsible
environment and humans with disastrous conse-
for changes in the Earth’s climate (https://www.bgs.
quences such as long-term shifts in global weather
ac.uk/discovering-geology/climate-change/what-
patterns and temperatures (https://news.un.org/en/
causes-the-earths-climate-to-change/). On the other
story/2022/05/1117842). In fact, the whole world is
hand, the increased water stress, a rise in the fre-
waging a war against nature for which man himself
quency of extreme events such as heatwaves, floods,
is responsible. Ever since the industrial revolution
cyclones, snowstorms, poor agricultural productivity,
started about 150 years back, human activities espe-
drought conditions, extinction of different forms of
cially related to power generation and various other
fauna and flora, spreading of new diseases, etc., are
industries, transportation, and agriculture have indis-
the key indicators of climate change (https://climate.
criminately released huge amounts of major green-
ec.europa.eu/climate-change/consequences-cli-
house gases (GHGs) such as carbon dioxide (CO2),
mate-change_en). Climate-related studies indicated
methane (CH 4), and nitrous oxide (N 2O) into the
that increased GHGs concentrations, land and ocean
environment disrupting the natural phenomenon and
surface temperatures, sea level rise, and ocean acid-
the delicate ecosystem of the planet (https://www.
epa.gov/ghgemissions/overview-greenhouse-gases). ification are the key indicators of climate change
Concentrations of these GHGs are currently more during the past few decades (Figure 1). The Inter-
abundant in the earth’s atmosphere than at any time governmental Panel on Climate Change (IPCC) is a
in the last 800,000 years. In 2021 our planet Earth United Nations body set up in 1988 to assess the sci-
reached a dark milestone of hitting over 150% of ence around climate change, and to provide govern-
CO2 value in the atmosphere compared to that of ments with scientific information which can be used
the pre-industrial times (https://www.noaa.gov/ to develop policies on reducing global warming.
news-release/carbon-dioxide-now-more-than-50- IPCC’s 6th Assessment Report (AR6), which details
higher-than-pre-industrial-levels). According to the the devastating consequences of GHG emissions
National Oceanic and Atmospheric Administration around the world, was released in March 2023. It
(NOAA), the average surface temperature across also identifies available, and cost-effective solutions
land and ocean in 2021 was 0.86 °C warmer than the to reduce GHG emissions to secure a safe, liveable
twentieth-century average of 13.9 °C, and 1.06 °C future. The average increase in global temperatures
warmer than the pre-industrial period (https://www. is now about 1.1 °C above pre-industrial levels ac-
climate.gov/news-features/understanding-climate/ cording to the IPCC (2023). To safeguard the Paris
climate-change-global-temperature). In fact, the Agreement, there is a need to limit global warming
experiments conducted by the scientific community to well below 2 °C, and if necessary drastic actions
even in the 1800s gave hints of human-induced are required to pursue efforts to limit it to 1.5 °C.
climate change as CO2 and other gases are capable For example, the Himalayan ecosystem is very del-
of altering the earth’s climate. But by the 1950s, the icate and even the slightest changes in the climate
subsequently generated data supported the global can have long-lasting effects on such vulnerable re-
warming theory [1]. gions [2]. The average global temperature across the
Carbon dioxide and other GHGs were found to be year was around 14.9 °C, putting it around 1.2 °C
responsible for the increase in the average tempera- rise above the average (Figure 2). The planet Earth

2
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 1. Natural and human factors are responsible for the earth’s climate, and key indicators.

Source: Modified after U.S. Global Change Research Program, Fourth National Climate Assessment, Chapter 2: Our Changing Climate, 2017.

ical and magnetic studies of the sediment [4]. Human

activities such as industries, electric power genera-
tion by coal, agriculture, and transportation are the
four top sources of GHGs which are responsible for
almost all of the increase in greenhouse gases in the
atmosphere over the last 150 years [5]. More pieces
of evidence are seen in recent times about human-in-
duced climate change [6]. Deforestation, agriculture,
and the use of fossil fuels are the primary sources of
CO2. China, India, the US, Russia, and Japan are the
biggest emitters of CO2 in the world today.

Figure 2. The increase in Earth’s average surface temperature

with time (from Wikipedia, 2023).
2. Greenhouse effect
Between the year 1870 (the first industrial revolu-
is neither too close nor too far away from the Sun tion) and today, the increase in global temperature is
and lies in a “Goldilocks zone” which is just right due to more fossil fuel burning (petroleum oil, nat-
to support life—not too hot, not too cold with the ural gas, coal), which also increased the CO2 levels
planet’s average surface temperature at 14.9 °C [3]. from 280 ppm in preindustrial time to 423.46 ppm in
But currently, with remarkable changes in our cli- June 2023 (NOAA, 2023 & https://www.co2.earth/
mate, a lot of questions are arising. A recent study daily-co2). The burning of fossil fuels and indiscrim-
on a 300-foot-long core of mud from a lakebed high inate cutting of forests increases CO2 in the atmos-
in the Peruvian Andes was found to preserve the phere (15-35 km above the Earth’s surface), and the
history of the deformation and flow of glaciers for increase in CO2 causes more heat to be trapped in the
more than 700,000 years. This is one of the longest Earth’s atmosphere as CO2 has the ability to absorb
records of historical climate revealing the variations infrared radiation. As a result, global temperatures
in temperature and precipitation based on geochem- are rising. The Earth receives radiation during the

3
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

daytime from the Sun which is a flow of infrared ria-almost-impossible-without-climate-change/).

radiation. Part of this is reflected back into space
by the atmosphere, while the rest warms the planet 3. Rise in methane gas in the envi-
during the day and is radiated back into space at ronment
night (https://earthobservatory.nasa.gov/features/En-
ergyBalance). During this delicate balance that pre- The latest report from the IPCC leaves no doubt
vailed in preindustrial times, the incoming heat was about the rising levels of CO2 and the powerful link
exactly offset by the amount lost to space, ensuring between methane (CH4) and climate change has also
average global temperatures remained constant. But been established already. Methane with more heat
unfortunately, today, due to the drastic increase in absorbing capacity has more warming potential and
GHGs, this balance is lost because some of the heat is 84 times stronger than CO2 on a 20-year horizon [8].
that should be radiated back into space gets trapped Emissions of heat-trapping methane are rising all
inside the atmosphere. Thus, the temperature is con- over the world, mostly due to agriculture, and intense
stantly rising as the concentrations of CO2 which has fossil fuel use [9,10], with the current concentration of
about 82% share among greenhouse gases, along methane reporting at 1909 ppb against a pre-indus-
with other greenhouse gases such as CH4, N2O, and trial time concentration of 722 ppb. The IPCC also
O3, increase in the atmosphere. This is known as the highlighted the problem posed by methane, which
greenhouse effect and these gases are trapping the in- has contributed as much as 0.5 °C of warming since
frared radiation and making the Earth warmer which preindustrial times, second only to CO2 [11]. Last
is also called global warming [7]. In order to combat year, about 111 countries that were responsible for
this phenomenon and to prevent the worst effects of nearly half of the methane emissions promised to cut
climate change, humans must reduce the release of down methane emissions by 2030 at COP26 in 2021
CO2 and other GHGs into the atmosphere and this (https://www.iea.org/reports/global-methane-track-
is the only option available for humanity. The ris- er-2022/the-global-methane-pledge). Methane usu-
ing temperatures could have catastrophic effects on ally leaks from oil and gas pipelines during main-
ecosystems if these trends continue to persist. The tenance operations which is evidenced by the huge
rapid increase in temperature would wipe out species methane plumes from oil and gas seen on satellite
not able to adapt quickly enough to this warming maps. It is also produced by landfill, agriculture, and
environment which is already visible all across the in coal production. Zhou et al. [12] evaluated the scale
planet. Polar ice caps could melt raising the sea lev- of methane emissions from the U.S. ammonia ferti-
els and causing flooding of coastal cities. A number lizer industry and found that the majority of methane
of pieces of evidence are being gathered on how hu- gets into the atmosphere through leakage only. At
man interference in the Earth’s climate continues to the same time, the methane emissions from the in-
impact global warming. Europe, the western US, and dustrial sector have been vastly underestimated and
Canada lived through the warmest summer tempera- are significantly higher. Recently Lauvaux et al. [13]
ture records during the last 3-4 years due to extreme focused their studies on detecting oil and gas leaks
wildfires which burnt several towns to the ground that can be plugged if companies want to invest in
and killed hundreds (https://climate.ec.europa.eu/ prevention. In addition, about 2% of the atmospheric
climate-change/consequences-climate-change_ methane is from the seafloor methane seepage due to
en). In 2023 some of the European countries like the gas hydrate dissociation activity contributed by
Spain recorded the highest temperature of the sea- climate change [14,15]. Despite these recent studies, an
son (https://www.worldweatherattribution.org/ appropriate understanding of the interplay between
extreme-april-heat-in-spain-portugal-morocco-alge- methane sources and sinks is not completely known.

4
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

4. Where from GHGs come and

where do they go?
GHGs including CO2 come from almost every
human activity such as the burning of fossil fuels
for transportation, electricity generation and run-
ning several industries, agriculture, land-use, and
residential and commercial activities. In 2014, the
top CO 2 emitters were China, the United States,
the European Union, India, the Russian Federation, Figure 3. Approximate global GHG emissions from different
and Japan (https://www.epa.gov/ghgemissions/ sources in the US. Total emissions are 6,558 million metric tons
global-greenhouse-gas-emissions-data). Apart from of CO2 equivalent.

coal-powered thermal power stations, transport, and Source: Modified after US EPA, 2019.

several other industries, volcanic eruptions all over

the world are some of the major sources of CO 2 ture’s fury are visible in many forms such as higher
(https://www.climate.gov/news-features/climate-qa/ average temperatures especially in colder regions,
which-emits-more-carbon-dioxide-volcanoes-or-hu- hotter summers even in regions such as Siberia,
man-activities). Over the geological times, before the melting sea ice in places like Arctic and Antarctic
pre-industrial times, variations in atmospheric CO2 regions, severe droughts, more wildfires, more pow-
depended mainly on volcanic emissions, which were erful hurricanes, and stronger winter storms. These
difficult to estimate. For example, the Etna volcano frequent and more intense extreme events are already
in Italy releases an exceptional amount of CO2 (9083 signaling our planet’s uncertain future. As a result,
t/day) and contributes to 10% of global volcanic currently, millions of lives, and homes are destroyed
emissions [16]. But CH4 and N2O, mostly come from in droughts and floods, and millions of people are
agriculture, and waste, and also from industries. For dislocated and facing hunger in certain parts of the
example, adipic acid is used in the production of world as more than 3 billion people are living in
industrial products such as nylon and polyurethanes highly vulnerable areas. These developments war-
and it is one of the largest sources of N2O emis- rant more serious investigations. New data show,
sions [17]. China alone accounts for 94% of global that the number of flooding and other hydrological
annual N2O emissions. Major contributors to GHG events all over the world has quadrupled since 1980
emissions in the USA in 2019 are shown in Figure and doubled since 2004 the world over [18]. Goswa-
3. Only about 50% of the CO2 from industrial emis- mi et al. [19] reported that there is a significant rise
sions remains in the atmosphere. The remainder is in the trend of extreme rain events and extreme heat
approximately equally divided between uptake into waves all over India in a warming environment. Ex-
the land biosphere and into the ocean. Most of the treme heat waves and wildfires are wreaking havoc
CO2 is absorbed by the plants, trees, minerals, and around the globe every year, and these wildfires usu-
microbes in the soils on the land, and oceans absorb ally send plumes of smoke into the higher reaches
a major portion, and the rest remains in the atmos- of our atmosphere, where smoke particles interact
phere (https://www.energy.gov/science/doe-explains- with ozone. There is growing evidence that bushfire
the-carbon-cycle). smoke reduces the concentration of ozone in the
stratosphere (20-25 km from the Earth’s surface),
as the tiny carbon particles originating from these
5. Increase in the frequency of ex- fires carry some chemical compounds containing
treme events chlorine which react with ozone, depleting its con-
The devastating effects of climate change and na- centration [20]. Ozone is highly effective at blocking

5
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

damaging ultraviolet rays from the sun. Without this the country in 2019 at 50.8 ºC, nine degrees above
layer, many plants would die, while humans and oth- normal. In addition, the number of extremely hot
er animals would be afflicted with skin cancers [21]. days every year when the temperature reaches 50 ºC
But it also contributes to the trapping of heat in the has doubled since the 1980s. India recorded its high-
troposphere. The drought-hit countries like Kenya, est temperature in February in 2023 with an average
Somalia, and Ethiopia in Africa have seen thousands maximum temperature was 29.54 ºC since recording
of families being displaced from their homes whilst began in 1901. Several other countries particularly
livestock losses have been huge. A small shift in the in Asia and Europe are currently experiencing very
center means that more of the curve touches the ex- high temperatures and severe dry conditions as rivers
tremes (Figure 4) and so heatwaves become more and freshwater lakes are getting dried up affecting
frequent and extreme as experienced, for example, their hydropower generation and agriculture leading
in western Canada and the US in 2021. In recent to severe drought and global food crisis. Millions of
years, the world has witnessed numerous wildfires fish are dying in some rivers in Europe and Australia
in countries such as Turkey, Greece, Italy, Spain, due to ongoing heatwaves in addition to toxic algal
Portugal, and France. The UK recorded over 40 blooms. Hotter temperatures are speeding up evapo-
ºC for the first time in history in 2022. In the usual ration leading to increased salt levels in the surface
weather cycle, hot weather creates more moisture water. Warmer water holds less oxygen than cold wa-
and the water vapor in the air turns into droplets to ter leading to hypoxia conditions as fish require more
create rain. The warmer it becomes, however, the oxygen in warmer conditions for their survival [25].
more vapor there is in the atmosphere, resulting in Sometimes the countries which are facing water
more droplets and heavier rainfall, sometimes in a stress are often accompanied by additional exposure
shorter space of time and over smaller areas such as to flooding (like India, and Pakistan) which have wit-
experienced by the historic flooding in certain parts nessed devastating rains and flooding with one-third
of China, Germany, Belgium, and the Netherlands in of Pakistan under water in 2022 [26]. IPCC (2023)
2021, and Pakistan in 2022. Cyclones in ice-covered report says that the frequency and the intensity of
areas such as Greenland and Russia in the Arctic those extreme events being witnessed these days are
which can destroy sea ice and threaten people living a result of global warming which will become more
in these regions are becoming more intense and frequent and severe in the future as the planet contin-
frequent [22]. In fact, the polar regions and countries ues to warm.
such as Britain, Spain, and Italy in Europe were hit
hardest by global warming in 2022. Apart from Eu-
rope, there were significant heat record events in the
Middle East, Central Asia, China, Pakistan, and parts
of India. A comprehensive worldwide assessment
of heatwaves down to regional levels has revealed
that heatwaves have been increasing in frequency
and duration since the 1950s in nearly every part
of the world [23]. The extreme meteorological sur- Figure 4. A small shift in climate can make a big difference in
face air temperatures recorded to date are –89.2 ºC in the weather conditions (US EPA).
Antarctica, and 56.7 ºC in Death Valley, California [24].
Australia recorded the hottest day in 62 years as the
temperature rose to 50.7 ºC in the coastal town of
6. Glaciers are shrinking and vanishing
Onslow Western Australia in January 2022. Churu in Northern Alaska, Canada, Northern Europe,
Rajasthan, India recorded the highest temperature in Northern Russia, Himalayas are melting faster at

6
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

the glacier bottom, shortening the glacier and reduc- entire coastlines in different parts of the world have
ing its mass, and the melt line moves upwards. A been a primary scientific focus in climate change
team of researchers [27] extracted ice cores from the studies for the past two decades as sea level rise is
glaciers on Mt. Hunter, in Alaska, and showed that one of the primary indicators. The melting of global
the amount of water that melts currently is 60 times ice sheets is one of the primary causes of sea lev-
greater than it was prior to 1850. These studies are el rise, which has accelerated over recent decades.
revealing that global temperatures are rising and in Since 1993, the global sea level has risen 4.3 cm (1.7
places like Greenland, the ice is vanishing and the inches), 60% coming from the expansion of ocean
destruction of coral reefs and other climate-related water as its temperature rise, and 40% from melting
damages are hitting the world at an unprecedented glaciers. In coastal areas where 40% of the world’s
rate. Satellite monitoring studies revealed that the population lives, rising sea levels can increase the
overall rate of ice loss from Western Antarctica has risk of flooding and endanger water sources. Pacific
increased five-fold over a 25-year period. It is inter- Island nations are already facing this threat, where
esting to note that Northern Hemisphere lost more people are being continuously displaced from their
ice (58%) in the past compared to that (42%) of the homes. Parts of cities such as Shanghai and Mumbai
Southern Hemisphere [28]. Relatively a smaller num- which are located in coastal areas are already experi-
ber of studies were carried out on Himalayan glaciers encing flooding due to global warming, and the rise
due to problems with accessing remote and high-el- in sea levels along the coast (https://climate-adapt.
evation glaciers, in politically unstable regions such eea.europa.eu/en/metadata/publications/ranking-
as Nepal, Pakistan, China, and India, making these of-the-worlds-cities-to-coastal-flooding/11240357).
mountains difficult places to work especially for for- According to some studies, the global sea level is not
eign scientists. In fact, Himalayan glaciers, are also rising steadily, rather it is getting faster every year.
showing increasing extreme temperatures and pre- The global-mean sea level reconstruction shows a
cipitation events, accelerating glacier-ice mass loss, trend of 1.5 ± 0.2 mm/year from 1958 to 2014 [32].
permafrost degradation, frequent landslides, and a
Globally, the sea level has risen about eight inches
substantial shift in the seasonal riverine water sup-
(20 cm) since the beginning of the 20th century and
ply. Terrestrial water storage and glacier show de-
more than two inches (5 cm) in the last 20 years
creasing trends in most mountain regions across the
alone. IPCC projections for 2100 range from 9 to 88
planet and will have a greater influence on glaciers,
cm. Antarctica is covered by an up to 4 km thick ice
groundwater, soil water, reservoirs, and lakes [29].
layer and contains about 90% of fresh water on the
In addition, glacial lake outburst floods represent a
surface of the Earth. This ice sheet in Antarctica is
major hazard in countries such as India, Pakistan, Peru
even shrinking in response to climate change and has
and China where over 15 million people live in high
lost over 3,000 billion tons of ice over 25 years [33].
mountain areas and are exposed to these risks [30]. Shuk-
Global sea level rise is 3 mm per year which is
la and Sen [31], recommend satellite-based, real-time
quite significant [34]. The rise in sea level not only
monitoring to develop a strong network of early
encroaches on territories of coastal communities but
flood warning systems in the Himalayas in order to
also contributes to seawater intrusion on freshwater
check the devastating flash floods which are occur-
systems in coastal areas. Coastal wetlands are critical
ring in higher regions of the Himalayan mountains
to Earth’s ecosystem which is becoming increasingly
more frequently in recent times.
vulnerable in the face of rising sea levels [6]. Recent-
ly, at the COP26 meeting in Glasgow, representatives
7. Sea levels rise and their impact on of some 1.2 billion people, from the Climate Vulner-
coastal zones able Forum (CVF) consisting of countries from Latin
Climate-induced sea-level rises recorded along America, Africa, Asia, the Caribbean, and the Pacific

7
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

were pushing the rest of the world to accept the idea ocean surface which affects the marine ecosystem [42].
of keeping the rise in global temperatures to under Ocean water’s acidity level becomes higher when
1.5 ºC. Currently, the world’s ocean surface heat is more CO2 dissolves in the water, although some CO2
at record-breaking levels and its impacts are poten- returns to the atmosphere from the oceans and land
tially catastrophic [35]. The vast amount of Arctic ice reservoirs due to the outgassing effect. This makes it
helps to cool the planet by reflecting the majority difficult for corals to build their skeleton [43] and cor-
of the Sun’s radiation back into space. But unfortu- al reefs are under constant threat as the atmospheric
nately, since the mid-1990s, there has been a marked CO 2 increases. In fact, both ocean warming and
decrease in sea ice cover in the Arctic Ocean [36] and acidification are two of the greatest global threats to
there is a fear among the scientific communities that the persistence of coral reefs [44]. In addition, if the
most of it could be gone in the coming 30 years [37]. International Seabed Authority (ISA) gives permis-
According to some estimates, rising seas could sub- sion to the mining of the seafloor for critical mineral
merge Jakarta (by 2100), Rio, and low-lying islands resources, the disturbance to the ocean’s ecosystems
like those of the Maldives, where half a million can also reduce their ability to absorb carbon to keep
people live barely a meter above the Indian Ocean, climate change under check [45].
are under threat of rising sea levels driven by global
warming. As a result, Indonesia is currently shifting 9. Water scarcity
its capital from Jakarta to the island of Borneo. The
Mediterranean Sea level rise by 2100 will lead to a Groundwater has been extensively exploited
potential loss of about 192 km2 of land along a coast- worldwide during the last several decades. As a re-
line length of about 400 km of locations of tourist re- sult, problems such as groundwater decline, and con-
sorts, railways, airports, and heritage sites belonging tamination are being currently confronted, and there
to France, Spain, Tunisia, and Cyprus [38]. Coming to is a definite threat to its sustainable use as a clean
the Indian Ocean, the sea level rise is seen along the water source in the coming future unless proper early
entire Indian coast and is rising faster than the global actions are taken. In addition, safeguarding ground-
average [39,40]. water quality is an essential part of sustainable water
management and taking care of human health [25,46].
8. Ocean acidification This deficiency of water is prevailing currently in
several countries such as Madagascar, Yemen, India,
Oceans cover over 70% of the Earth’s surface and and China, and in many European countries for sev-
play a crucial role in sequestering significant quan- eral years across the globe. This situation has posed
tities of CO2 from the atmosphere. The free-float- more serious questions about the disparity of rainfall
ing algae called phytoplankton and tiny planktonic caused by climate change and global warming, and
animals called zooplankton in oceans play a major over 4 billion people in the world are currently fac-
role in the absorption of CO2 and transfer of it to the ing water scarcity [47]. At present, the whole world is
ocean floor. The rise in CO2 levels has led to more suffering from the spread of new viral diseases such
absorption and thereby more acidification of the as COVID-19 and its several variants, monkeypox,
oceans, leading to the weakening of the shells and and tomato flu, threatening the health of people the
skeletons of animals living in the sea. Concentrations world over. According to Lim et al. [48], there is a
of elements such as Mo, B, V, Cr, U, and Se and threat of increased infectious disease risk likely in
their isotopes in seawater and marine animals allow the future.
an understanding of ancient ocean chemistry and the
ocean acidification process [41]. The replenishment
of fish populations is threatened by ocean acidifica-
10. Effect on food production
tion caused by the uptake of additional CO2 at the As the global land surface temperature is on the

8
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

rise, our key food and fuel cropping systems are With favorable conditions such as a low population,
expected to affect our food production. In general, large amounts of good quality agricultural land, and
the increase in CO2 concentrations together with the reliable domestic energy, New Zealand is expected
increase in global temperatures theoretically produc- to hold up the best, compared to others. So, even if
es greater yields in crops. However, the majority of climate change triggers a catastrophic global civili-
studies have shown that crop yields are downward zation collapse, humans will likely be able to keep
trending due to the increase in the frequency of going, at least, in some parts of the world. Coming
extreme weather events [49]. At temperatures above to animals, polar bears are some of the largest terres-
40 °C, the photosynthesis rate slows down because trial carnivores on Earth, but their fate is intimately
the enzymes involved in the chemical reactions of tied to what happens to the Arctic’s sea ice. Most
photosynthesis are temperature sensitive and get polar bears are expected to struggle to survive in the
destroyed at greater warming conditions. Excessive Arctic by 2100 as the Arctic is continuously losing
heat also can reduce the efficiency of enzymes that its ice. Standing almost four feet tall and weighing
drive photosynthesis and can hinder plants’ ability to around 40 kg, emperor penguins in Antarctica are the
regulate CO2 uptake and water loss [50]. Hence with world’s largest penguins which need sea ice for their
more and more global warming in the future, there is survival. A recent study estimates that 98% of colo-
a definite threat to food production worldwide. nies will be quasi-extinct by 2100 unless the world
drastically reduces its greenhouse gas emissions [54].
11. The sixth mass extinction? Bumble bees which depend on wildflowers for their
survival can also be affected as extremely high tem-
There have been five mass biodiversity extinction peratures have negatively impacted most species of
events in the 4.54 billion years of the history of the them over the past 120 years [55]. If greenhouse gas
Earth, all caused by dramatic and extreme natural emissions and warmer temperatures are not con-
phenomena such as prolonged ice age, sea-level trolled, some researchers estimate that life in oceans
changes, asteroid impacts, and voluminous volcanic could be depleted severely within the coming 150
eruptions in certain parts of the world. The massive years [56]. The risk of species loss depends on how far
volcanic eruptions leading to the formation of the Si- global temperatures rise in the future.
berian Traps, and the enormous amounts of CO2 led
to the wiping out of 90% of all marine life during the
largest extinction event (Permian-Triassic) in Earth’s
12. Current, and upcoming technol-
history [51]. Today, there is mounting evidence that
ogies to limit warming, and steps to
the Earth is at the start of, or perhaps in the midst of
net zero
the sixth mass extinction crisis which is underway Cutting down CO2 emissions to control climate
with 7.5%-13% of species lost over the last 500 change and reach net zero by 2050 is possible but
years, entirely caused by human activities [52]. Even not easy. So far, we have failed to reverse the 150-
if such an eventuality is unavoidable, some countries year trend of rising greenhouse gas emissions even
can most likely survive a global societal collapse after repeated warnings by the scientific communi-
and maintain their complex way of life according ties for more than 30 years (IPCC, 2023). The UN
to a study by King and Jones [53]. For this study, five secretary general is calling for countries to bring
island nations, including Ireland and New Zealand forward their plans for the effective reduction of
were chosen as they could remain habitable through GHG emissions for net zero by a decade or at least
agriculture because of their vast land area for a small as close as possible to 2040. Innovative technolo-
group of people they have, relatively cool temper- gies such as solar-driven hydrogen production and
atures, low weather variability, and other factors fuel cell technologies [57,58], and space-solar power
that make them more resistant to climate change. are likely to help to achieve sustainable goals and

9
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

for environmental planning by fulfilling the require- by different international firms during the last couple
ment of energy if they are technically feasible [59]. of years. Unfortunately, the implementation of these
NASA and European Space Agency (ESA) are of and a few other critical and proven technologies has
the opinion that space-based solar power technology been very slow and inadequate [64]. Geological CCS
has the ability to increase energy independence and technology involves the injection of CO 2 under-
reduce greenhouse gas emissions. In addition, sev- ground for long-term geological storage. It can help
eral countries have recognized the potential of wave fight climate change by lowering industrial emis-
power that could tackle the global energy crisis, sions [65]. Carbon dioxide would be injected into the
support economic growth, and provide green ener- ground for sequestration and geological medium like
gy [60]. Several studies are also currently focused on igneous and metamorphic rocks through a process
more efficient solar panels reaching efficiency levels known as carbon mineralization are needed for keeping
nearing 50%. Four pioneering scientists who built the CO2 underground [66] (Figure 5). In this context, ol-
more efficient electricity-generating silicon solar ivine mineral is found to be very efficient in absorbing
cells (from around 16-18% efficiency to about 25%) CO2 and this process is also up-scalable [67].
have won Queen Elizabeth Prize for Engineering in
2023. This is an example of the intensive efforts to-
wards the development of technologies for tackling
climate change, and the recognition and encourage-
ment of the technologists and scientists who achieve
stunning successes in this direction. The estimates
of some scientists suggest that by the 2030s, solar
will have more installed capacity than coal, oil, gas,
nuclear, and hydro put together. Switching over to
battery-powered electric vehicles or vehicles based
on hydrogen fuel or using hydrogen fuel cells in the
coming years will certainly help to reduce CO2 and
tackle climate change, despite the fact that there are
also large uncertainties around the emissions associ-
ated with electric vehicle battery production. How-
ever, according to Plötz [61], fuel-cell electric vehicle
technology will not play a major role in sustainable
road transport in the future. The CO2 concentration
of about 425 ppm poses a difficult situation for high Figure 5. Concept of carbon capture and storage.
CO2 capture capacities using sorption-desorption Source: Modified after Schiermeier et al. [68].
processes [62]. The magnitude of this crisis is gigantic,
with about 40 gigatons/year of global CO2 emission Net zero emissions target means reducing carbon
with no signs of shrinking [63]. Out of several tech- emissions to the greatest extent possible, by phas-
nologies, technologies like direct air capture (DAC) ing out coal-powered thermal stations, switching
which uses a 10% solution of KOH to capture CO2 over to alternate renewable energies such as hydro,
from the atmosphere, and carbon capture and storage solar, nuclear, hydrogen, and wave, and removing
(CCS) have emerged as indispensable tools in com- the remaining unavoidable emissions via removal
bating climate change and reaching their goal of net technologies. Planting more and more trees is seen
zero emissions. A lot of meetings and summits are as a good option to tackle climate change. Replac-
being organized to discuss the latest developments ing conventional fossil fuel vehicles with electric

10
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

vehicles which run on batteries would also signifi- for a sustainable energy future. Since land-based
cantly reduce CO2 in the atmosphere improving air mineral deposits are dwindling fast, sea-bed resourc-
quality and human health [69]. Every nation on Earth es are seen as a new resource frontier for mineral
understood the science behind climate change and exploration and extraction [73-75]. As the land ore de-
the catastrophic consequences the people have to posits are becoming scarce for some of these critical
face if man does not act promptly to keep the rise metals, there is an urgent need to look for alternative
in global temperatures well below 2 ºC by the end resources [76]. As vast mineral wealth is available in
of this century and pursue efforts to keep it under the seabed, the demand for exploiting critical min-
1.5 ºC. Several countries in Europe including the erals in the deep sea is rising in recent times. But
UK have already cut down the use of coal. India has what about the ecological and environmental conse-
set a target to ensure that 50% of its energy will be quences of mining the ocean which covers 71% of
from renewable energy sources by 2030 and also to the planet’s surface, and in turn is the habitat of the
achieve net zero emissions latest by 2070 as per the vast majority (97%) of living organisms? [45,77]. The
declaration at the COP26 summit in Glasgow last oceans are an essential part of the biosphere, influ-
year. In the meantime, a lot of studies are going on encing several areas such as climate, food, health,
with regard to the reduction of CO2 and converting and overall well-being.
it into useful products such as fuels, and chemicals
which also contribute to the reduction of CO2 levels 13. United and rigorous global ef-
and avoid catastrophic temperature rises across the forts only can resolve this crisis!
world [70]. The easiest way to cut down the release of
carbon is by using alternative ways to generate elec- Many nations are not able to achieve their sus-
tricity such as the use of hydropower, nuclear fission, tainable development goals such as poverty allevi-
fusion (future), biomass, wind, geothermal, solar, hy- ation, quality healthcare, and economic growth due
drogen, and ocean energy [69,71]. Already some coun- to climate change-driven lethal heatwaves and other
tries such as the UK are producing more electricity climate change-driven extreme events [78]. Unfortu-
for the first time by using renewable sources such nately, each country takes a different track toward
as wind power. At present, energy-efficient lighting the efforts for the mitigation of climate change. The
is saving up to 75% of the energy currently used big oil companies obtain approvals and licenses with
to light our homes, thanks to the invention of LED the promise that they store huge amounts of CO2
bulbs. In 2022, net fusion power was achieved for underground by CCS technology but fail to reach the
the first time, and some countries like the US, EU, promised storage goals. The EU targets to cut down
UK, and Japan are heavily investing in nuclear-fu- all greenhouse gases by 2050. China’s net-zero plan
sion power which is clean power and a breakthrough focuses on balancing CO2 emissions by 2060. The
can be expected at any time in the near future [72]. UK announced that it would reach net zero green-
house-gas emissions by 2050. Top oil exporter Saudi
Arabia targets net zero by 2060, and Australia pledg-
Supply of critical metals required for green tech-
es to reach net zero emissions by 2050. But without
nology applications
more clear strategies by different countries behind
For the whole world to deliver net zero by 2050, achieving net zero targets, it’s very difficult to evalu-
large-scale mining is more critical for metals such ate the impact. However, according to Rogelj et al. [11]
as Li, Cu, Ni, Co, Pt, Pd, REE, Ga, W, Te, and In, as the current pledges and actions by different govern-
these metals are essential for green technology appli- ments together with technological advancements,
cations such as making wind turbines, solar panels, the world will be able to witness the decline of GHG
fuel-cells, electric vehicles, and data storage systems only after 2050 through united and rigorous global
required to transition to a low-carbon economy and efforts (Figure 6).

11
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 6. Estimated global temperature peaks (in pink) and declines (arrows) under net-zero GHG emissions.

Source: Modified after [11].

14. Conclusions and future very optimistic about the restoration of the natural
climatic conditions as targeted, just as the notable
At the recent 26th United Nations Climate recovery of the protective ozone layer in the upper
Change Conference of the Parties (COP26), over 127 stratosphere has been observed in recent times due
countries committed to net zero targets which will to the complete ban on ozone-depleting substances.
limit global warming to 1.5 ºC above pre-industrial The UN Secretary-General António Guterres also
levels. Several scientists believe that the Paris said that the restoration of the ozone layer was “an
Climate Agreement’s goal to limit global warming encouraging example of what the world can achieve
in this century to 2 ºC is certainly within reach. All when we work together”.
governmental bodies are making climate emergency
declarations. It is possible to reach net zero carbon
Conflict of Interest
emissions, but it is not going to be easy and there
is a long road ahead to net zero. With available There is no conflict of interest.
emission budgets defined by the 2 ºC target itself
looked challenging, the goal of achieving 1.5 ºC Acknowledgement
looks certainly harder unless united and rigorous
The author thanks Dr. Prakash Kumar, Director
global efforts are made. A coordinated effort of all
of CSIR-NGRI, Hyderabad, for his support and en-
nations is required to switch over from fossil fuels
couragement. The contents of this manuscript were
to renewable energy systems such as solar, wind,
presented in part by the author at > 70 INSPIRE Sci-
and nuclear for electricity generation (Figure 6).
ence Camps organized by the Department of Science
The necessary technologies have to be developed
and Technology (DST), New Delhi, Government of
to remove excess CO2 and other GHGs, and partic-
India, in different Universities and Academic Insti-
ulate pollution from the atmosphere, and limit CO2
tutions across the country, during the last ~15 years,
emissions by planting more and more trees. Every
and also at the Trent Lott Geospatial and Visualiza-
individual must understand the severity of this situ-
tion Research Center, and the Department of Chem-
ation and contribute toward tackling climate change.
istry, Jackson State University, Jackson, MS, USA in
Children should be taught the cause, effects, and
2010.
how to combat climate change in depth even from
high school. Any further delay in concerted global
action will miss a brief and rapidly closing window
References
to secure a liveable future. In fact, all of us can be [1] Le Treut, H., Somerville, R., Cubasch, U., et al.,

12
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

2007. Historical overview of climate change. sions arise equally from agricultural and fossil
Climate change 2007: The physical science fuel sources. Environmental Research Letters.
basis. Contribution of Working Group I to the 15(7), 071002.
Fourth Assessment Report of the Intergovern- DOI: https://doi.org/10.1088/1748-9326/ab9ed2
mental Panel on Climate Change. Cambridge [10] Cheng, C.H., Redfern, S.A., 2022. Impact of
University Press: Cambridge, United Kingdom interannual and multidecadal trends on meth-
and New York, NY, USA. ane-climate feedbacks and sensitivity. Nature
[2] Kc, K.B., Tzadok, E., Pant, L., 2022. Himalayan Communications. 13(1), 3592.
ecosystem services and climate change driven DOI: https://doi.org/10.1038/s41467-022-31345-w
agricultural frontiers: A scoping review. Discov- [11] Rogelj, J., Geden, O., Cowie, A., et al., 2021.
er Sustainability. 3(1), 35. Net-zero emissions targets are vague: Three
DOI: https://doi.org/10.1007/s43621-022-00103-9 ways to fix. Nature. 591(7850), 365-368.
[3] Isokangas, E., Rozanski, K., Rossi, P.M., et al., DOI: https://doi.org/10.1038/d41586-021-00662-3
2015. Quantifying groundwater dependence [12] Zhou, X., Passow, F.H., Rudek, J., et al., 2019.
of a sub-polar lake cluster in Finland using an Estimation of methane emissions from the US
isotope mass balance approach. Hydrology and ammonia fertilizer industry using a mobile sens-
Earth System Sciences. 19(3), 1247-1262. ing approach. Elementa: Science of the Anthro-
[4] Rodbell, D.T., Hatfield, R.G., Abbott, M.B., et pocene. 7, 19.
al., 2022. 700,000 years of tropical Andean gla- DOI: https://doi.org/10.1525/elementa.358
ciation. Nature. 607(7918), 301-306. [13] Lauvaux, T., Giron, C., Mazzolini, M., et al.,
DOI: https://doi.org/10.1038/s41586-022-04873-0 2022. Global assessment of oil and gas methane
[5] Perera, F., 2018. Pollution from fossil-fuel com- ultra-emitters. Science. 375(6580), 557-561.
bustion is the leading environmental threat to DOI: https://doi.org/10.1126/science.abj4351
global pediatric health and equity: Solutions [14] Ruppel, C.D., Kessler, J.D., 2017. The interac-
exist. International Journal of Environmental tion of climate change and methane hydrates.
Research and Public Health. 15(1), 16. Reviews of Geophysics. 55(1), 126-168.
DOI: https://doi.org/10.3390/ijerph15010016 DOI: https://doi.org/10.1002/2016RG000534
[6] Weston, N.B., Rodriguez, E., Donnelly, B., et [15] Clemens, S.C., Thirumalai, K., Oppo, D., 2023.
al., 2023. Recent acceleration of wetland accre- Indian margin methane hydrate dissociation
tion and carbon accumulation along the US East recorded in the carbon isotopes of benthic (Mili-
Coast. Earth’s Future. 11(3), e2022EF003037. olida) foraminifera. Earth and Planetary Science
DOI: https://doi.org/10.1029/2022EF003037 Letters. 609, 118101.
[7] Gadani, H., Vyas, A., 2011. Anesthetic gases DOI: https://doi.org/10.1016/j.epsl.2023.118101
and global warming: Potentials, prevention and [16] Bragagni, A., Mastroianni, F., Münker, C., et
future of anesthesia. Anesthesia, Essays and Re- al., 2022. A carbon-rich lithospheric mantle as a
searches. 5(1), 5. source for the large CO2 emissions of Etna vol-
DOI: https://doi.org/10.4103/0259-1162.84171 cano (Italy). Geology. 50(4), 486-490.
[8] Zhang, Z., Poulter, B., Feldman, A.F., et al., DOI: https://doi.org/10.1130/G49510.1
2023. Recent intensification of wetland methane [17] Wu, T., Niu, Z., Feng, L., et al., 2021. Perfor-
feedback. Nature Climate Change. 13(5), 430- mance analysis of VPSA process for separating
433. N2O from adipic acid tail gas. Separation and
DOI: https://doi.org/10.1038/s41558-023-01629-0 Purification Technology. 256, 117750.
[9] Jackson, R.B., Saunois, M., Bousquet, P., et al., DOI: https://doi.org/10.1016/j.seppur.2020.117750
2020. Increasing anthropogenic methane emis- [18] Cai, W., Wang, G., Santoso, A., et al., 2015.

13
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Increased frequency of extreme La Niña events DOI: https://doi.org/10.1007/s10584-021-03277-9

under greenhouse warming. Nature Climate [27] Winski, D., Osterberg, E., Kreutz, K., et al.,
Change. 5(2), 132-137. 2018. A 400‐year ice core melt layer record
DOI: https://doi.org/10.1038/nclimate2492 of summertime warming in the Alaska Range.
[19] Goswami, B.N., Venugopal, V., Sengupta, D., Journal of Geophysical Research: Atmospheres.
et al., 2006. Increasing trend of extreme rain 123(7), 3594-3611.
events over India in a warming environment. DOI: https://doi.org/10.1002/2017JD027539
Science. 314(5804), 1442-1445. [28] Slater, T., Lawrence, I.R., Otosaka, I.N., et al.,
DOI: https://doi.org/10.1126/science.1132027 2021. Earth’s ice imbalance. The Cryosphere.
[20] Solomon, S., Dube, K., Stone, K., et al., 2022. 15(1), 233-246.
On the stratospheric chemistry of midlatitude DOI: https://doi.org/10.5194/tc-15-233-2021
wildfire smoke. Proceedings of the National [29] Li, X., Wang, L., Hu, B., et al., 2023. Contri-
Academy of Sciences. 119(10), e2117325119. bution of vanishing mountain glaciers to global
DOI: https://doi.org/10.1073/pnas.2117325119 and regional terrestrial water storage changes.
[21] Lucas, R.M., Norval, M., Neale, R.E., et al., Frontiers in Earth Science. 11, 1134910.
2015. The consequences for human health of DOI: https://doi.org/10.3389/feart.2023.1134910
stratospheric ozone depletion in association with [30] Taylor, C., Robinson, T.R., Dunning, S., et al.,
other environmental factors. Photochemical & 2023. Glacial lake outburst floods threaten mil-
Photobiological Sciences. 14(1), 53-87. lions globally. Nature Communications. 14, 487.
DOI: https://doi.org/10.1039/c4pp90033b DOI: https://doi.org/10.1038/s41467-023-36033-x
[22] Zhang, X., Tang, H., Zhang, J., et al. (editors), [31] Shukla, T., Sen, I.S., 2021. Preparing for floods
2022. Continuing intensification of arctic cy- on the Third Pole: Satellite-based real-time
clone activity. AGU Fall Meeting; 2022 Dec 12- monitoring is needed for Himalayan glacial
16; Chicago, IL. catchments. Science. 372(6539), 232-234.
[23] Perkins-Kirkpatrick, S.E., Lewis, S.C., 2020. DOI: https://doi.org/10.1126/science.abh3558
Increasing trends in regional heat waves. Nature [32] Frederikse, T., Jevreva, S., Riva, R.E.M., et al.,
Communications. 11, 3357. 2018. Consistent sea-level reconstruction and
DOI: https://doi.org/10.1038/s41467-020-16970-7 its budget on basin and global scales over 1958-
[24] Andrew, C., 2017. Global temperature and life. 2014. American Meteorological Society. 31(3),
Principles of thermal ecology: Temperature, en- 1267-1280.
ergy, and life. Oxford University Press: Oxford. DOI: https://doi.org/10.1175/JCLI-D-17-0502.1
pp. 308-328. [33] Davison, B.J., Hogg, A.E., Rigby, R., et al.,
DOI: https://doi.org/10.1093/oso/9780199551668. 2023. Sea level rise from West Antarctic mass
003.0014 loss significantly modified by large snowfall
[25] Balaram, V., Copia, L., Kumar, U.S., et al., anomalies. Nature Communications. 14, 1479.
2023. Pollution of water resources, causes, ap- DOI: https://doi.org/10.1038/s41467-023-36990-3
plication of ICP-MS techniques in hydrological [34] Wilson, D.J., Bertram, R.A., Needham, E.F., et
studies, monitoring, and management. Geosys- al., 2018. Ice loss from the East Antarctic Ice
tems and Geoenvironment. 2(4), 100210. Sheet during late Pleistocene inter-glacials. Na-
DOI: https://doi.org/10.1016/j.geogeo.2023.100210 ture. 561, 383-386.
[26] Warren, R., Andrews, O., Brown, S., et al., DOI: https://doi.org/10.1038/s41586-018-0501-8
2022. Quantifying risks avoided by limiting [35] Cheng, L., Abraham, J., Trenberth, K.E., et al.,
global warming to 1.5 or 2 °C above pre‑indus- 2023. Another year of record heat for the oceans.
trial levels. Climatic Change. 172, 39. Advances in Atmospheric Sciences. 40(6), 963-

14
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

974. [44] Cornwall, C.E., Comeau, S., Putnam, H., et al.,

DOI: https://doi.org/10.1007/s00376-023-2385-2 2022. Impacts of ocean warming and acidifica-
[36] Ivanov, V., 2023. Arctic sea ice loss enhances tion on calcifying coral reef taxa: Mechanisms
the oceanic contribution to climate change. At- responsible and adaptive capacity. Emerging
mosphere. 14(2), 409. Topics in Life Sciences. 6(1), 1-9.
DOI: https://doi.org/10.3390/atmos14020409 DOI: https://doi.org/10.1042/ETLS20210226
[37] Hu, A., Meehl, G.A., Han, W., et al., 2009. Tran- [45] Balaram, V., 2023. Deep-sea mineral deposits
sient response of the MOC and climate to poten- as a future source of critical metals, and envi-
tial melting of the greenland ice sheet in the 21st ronmental issues—a brief review. Minerals and
century. Geophysical Research Letters. 36(10). Mineral Materials. 2(2), 5.
DOI: https://doi.org/10.1029/2009GL037998 DOI: https://doi.org/10.20517/mmm.2022.12
[38] Antonioli, F., Falco, G.D., Presti, V.L., et al., [46] Balaram, V., Rani, A., Rathore, D.P.S., 2022.
2020. Relative sea-level rise and potential sub- Uranium in groundwater in parts of India and
mersion risk for 2100 on 16 coastal plains of the world: A comprehensive review of sources, im-
Mediterranean Sea. Water. 12(8), 2173. pact to the environment and human health, ana-
DOI: https://doi.org/10.3390/w12082173 lytical techniques, and mitigation technologies.
[39] Unnikrishnan, A.S., Nidheesh, A.G., Lengaigne, Geosystems and Geoenvironment. 1(2), 100043.
M., 2015. Sea-level-rise trends off the Indian DOI: https://doi.org/10.1016/j.geogeo.2022.100043
coasts during the last two decades. Current Sci- [47] Mekonnen, M.M., Hoekstra, A.Y., 2016. Four
ence. 108(10), 966-971. billion people facing severe water scarcity. Sci-
[40] Krishnan, R., Sanjay, J., Gnanaseelan, C., et al., ence Advances. 2(2), e1500323.
2021. Assessment of climate change over the DOI: https://doi.org/10.1126/sciadv.1500323
Indian Region. A report of the Ministry of Earth [48] Lim, D.K., Kim, J.W., Kim, J.K., 2022. Effects
Sciences (MoES), Government of India. Spring- of climatic factors on the prevalence of influen-
er Nature: Switzerland AG. za virus infection in Cheonan, Korea. Environ-
DOI: https://doi.org/10.1007/978-981-15-4327-2 mental Science and Pollution Research. 29(39),
[41] Balaram, V., Rahaman, W., Roy, P., 2022. Re- 59052-59059.
cent advances in MC-ICP-MS applications in DOI: https://doi.org/10.1007/s11356-022-20070-y
Earth and environmental sciences: Challenges [49] Mirón, I.J., Linares, C., Díaz, J., 2023. The in-
and solutions. Geosystems and Geoenvironment. fluence of climate change on food production
1(2), 100019. and food safety. Environmental Research. 216,
DOI: https://doi.org/10.1016/j.geogeo.2021.100019 114674.
[42] Munday, P.L., Dixson, D.L., McCormick, M.I., DOI: https://doi.org/10.1016/j.envres.2022.114674
et al., 2010. Replenishment of fish populations [50] Moore, C.E., Hensold, K.M., Lemonnier, P., et
is threatened by ocean acidification. Proceedings al., 2021. The effect of increasing temperature
of the National Academy of Sciences. 107(29), on crop photosynthesis: From enzymes to eco-
12930-12934. systems. Journal of Experimental Botany. 72(8),
DOI: https://doi.org/10.1073/pnas.1004519107 2822-2844.
[43] Sulpis, O., Boudreau, B.P., Mucci, A., et al., DOI: https://doi.org/10.1093/jxb/erab090
2018. Current CaCO3 dissolution at the seafloor [51] Penn, J.L., 2022. Avoiding ocean mass extinc-
caused by anthropogenic CO2. Proceedings of tion from climate warming. Science. 376(6592),
the National Academy of Sciences. 115(46), 524-526.
11700-11705. DOI: https://doi.org/10.1126/science.abe9039
DOI: https://doi.org/10.1073/pnas.1804250115 [52] Cowie, R.H., Bouchet, P., Fontaine, B., 2022.

15
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

The sixth mass extinction: Fact, fiction or spec- DOI: https://doi.org/10.1016/j.rser.2021.110932

ulation? Biological Reviews. 97(2), 640-663. [61] Plötz, P., 2022. Hydrogen technology is unlikely
DOI: https://doi.org/10.1111/brv.12816 to play a major role in sustainable road trans-
[53] King, N., Jones, A., 2021. An analysis of the po- port. Nature Electronics. 5, 8-10.
tential for the formation of ‘nodes of persisting [62] Chen, H., Dong, H., Shi, Z., et al., 2023. Direct
complexity’. Sustainability. 13(15), 8161. air capture (DAC) and sequestration of CO 2:
DOI: https://doi.org/10.3390/su13158161 Dramatic effect of coordinated Cu (II) onto a
[54] Jenouvrier, S., Che‐Castaldo, J., Wolf, S., et al., chelating weak base ion exchanger. Science Ad-
2021. The call of the emperor penguin: Legal re- vances. 9(10), eadg1956.
sponses to species threatened by climate change. DOI: https://doi.org/10.1126/sciadv.adg1956
Global Change Biology. 27(20), 5008-5029. [63] Friedlingstein, P., O’Sullivan, M., Jones, M.W.,
DOI: https://doi.org/10.1111/gcb.15806 et al., 2022. Global carbon budget 2022. Earth
[55] Jackson, M.H., Johnson, S.A., Morandin, L.A., System Science Data. 14(11), 4811-4900.
et al., 2022. Climate change winners and losers DOI: https://doi.org/10.5194/essd-14-4811-2022
among North American bumblebees. Biology [64] Morrow, D.R., 2021. Is there a role for carbon
Letters. 18(6), 071002 capture and storage in a just transition? One
DOI: https://doi.org/10.1098/rsbl.2021.0551 Earth. 4(11), 1546-1547.
[56] Crichton, K.A., Wilson, J.D., Ridgwell, A., et DOI: https://doi.org/10.1016/j.oneear.2021.10.022
al., 2023. What the geological past can tell us [65] Meckel, T.A., Bump, A.P., Hovorka, S.D., et al.,
about the future of the ocean’s twilight zone. 2021. Carbon capture, utilization, and storage
Nature Communications. 14, 2376. hub development on the Gulf Coast. Greenhouse
DOI: https://doi.org/10.1038/s41467-023-37781-6 Gases: Science and Technology. 11(4), 619-632.
[57] Song, H., Luo, S., Huang, H., et al., 2022. So- DOI: https://doi.org/10.1002/ghg.2082
lar-driven hydrogen production: Recent advanc- [66] Shaw, R., Mukherjee, S., 2022. The develop-
es, challenges, and future perspectives. ACS ment of carbon capture and storage (CCS) in
Energy Letters. 7(3), 1043-1065. India: A critical review. Carbon Capture Science
DOI: https://doi.org/10.1021/acsenergylett.1c02591 & Technology. 2, 100036.
[58] Kastell, D., 2022. Hydrogen storage technology DOI: https://doi.org/10.1016/j.ccst.2022.100036
for aerial vehicles. Fuel cell and hydrogen tech- [67] Vink, J.P., Knops, P., 2023. Size-fractionated
nologies in aviation. Springer Nature: Switzer- weathering of olivine, its CO2-sequestration rate,
land AG. and ecotoxicological risk assessment of nickel
[59] Dakora, J.D., Davidson, I.E., Sharma, G. (ed- release. Minerals. 13(2), 235.
itors), 2020. Review of modern solar power DOI: https://doi.org/10.3390/min13020235
satellite and space rectenna systems. 2020 Inter- [68] Schiermeier, Q., Tollefson, J., Scully, T., et al.,
national Conference on Artificial Intelligence, 2008. Energy alternatives: Electricity without
Big Data, Computing and Data Communication carbon. Nature. 454, 816-823.
Systems (icABCD); 2020 Aug 6-7; Durban, DOI: https://doi.org/10.1038/454816a
South Africa. New York: IEEE. [69] Balaram, V., 2020. Environmental impact of Pt,
DOI: https://doi.org/10.1109/icabcd49160. Pd and Rh emissions from autocatalytic convert-
2020.9183884 ers—A brief review of the latest developments.
[60] Jin, S., Greaves, D., 2021. Wave energy in the Handbook of environmental materials manage-
UK: Status review and future perspectives. Re- ment. Springer Nature: Switzerland AG. pp.
newable and Sustainable Energy Reviews. 143, 1-37.
110932. [70] Yao, B., Xiao, T., Makgae, O.A., et al., 2020.

16
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Transforming carbon dioxide into jet fuel using Geoscience Frontiers. 10(4), 1285-1303.
an organic combustion-synthesized Fe-Mn-K DOI: https://doi.org/10.1016/j.gsf.2018.12.005
catalyst. Nature Communications. 11(1), 6395. [75] Balaram, V., 2023. Potential future alternative
DOI: https://doi.org/10.1038/s41467-020-20214-z resources for rare earth elements: Opportunities
[71] Yadav, K., Sircar, A., 2021. Geothermal energy and challenges. Minerals. 13(3), 425.
provinces in India: A renewable heritage. Inter- DOI: https://doi.org/10.3390/min13030425
national Journal of Geoheritage and Parks. 9(1), [76] Toro, N., Galvez, E., Saldana, M., et al., 2022.
93-107. Submarine mineral resources: A potential solu-
DOI: https://doi.org/10.1016/j.ijgeop.2020.12.002 tion to political conflicts and global warming.
[72] Betti, R., 2023. A milestone in fusion research is Minerals Engineering. 179, 107441.
reached. Nature Reviews Physics. 5(1), 6-8. DOI: https://doi.org/10.1016/j.mineng.2022.
DOI: https://doi.org/10.1038/s42254-022-00547-y 107441
[73] Hein, J.R., Mizell, K., Koschinsky, A., et al., [77] Turner, P.J., 2019. Deep-sea mining and envi-
2013. Deep-ocean mineral deposits as a source ronmental management. Encyclopedia of Ocean
of critical metals for high-and green-technology Sciences, 3rd Edition. 6, 507-515.
applications: Comparison with land-based re- DOI: https://doi.org/10.1016/B978-0-12-
sources. Ore Geology Reviews. 51, 1-14. 409548-9.11106-6
DOI: http://dx.doi.org/10.1016/j.oregeor- [78] Debnath, R., Bardhan, R., Bell, M.L., 2023.
ev.2012.12.001 Lethal heatwaves are challenging India’s sus-
[74] Balaram, V., 2019. Rare earth elements: A re- tainable development. PLOS Climate. 2(4),
view of applications, occurrence, exploration, e0000156.
analysis, recycling, and environmental impact. DOI: https://doi.org/10.1371/journal.pclm.0000156

17
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Journal of Geographical Research

https://journals.bilpubgroup.com/index.php/jgr

ARTICLE

Spatiotemporal Analysis of Land Use Land Cover Mapping and

Change Detection in Dambatta Local Government Area
David Sesugh Aule1*, Mamman Saba Jibril1, Ali Hussain Idris2
1
Department of Geography, Nigerian Defence Academy, Kaduna, PMB 2109, Nigeria
2
Nigerian Airforce Headquarters, Abuja, PMB 0244, Nigeria

ABSTRACT
This research studied the spatiotemporal changes in land use (LU)/land cover (LC) in Dambatta local government
area, with a view to identifying the effect arising from the observable changes in land use patterns. The imageries
used in the study were obtained from the National Space Research and Development Agency (NARSDA), Abuja.
Spatial analytical techniques and descriptive statistical techniques were employed to analyze the data. The results
showed 66.8% reduction in agricultural lands, 45.5% reduction in vegetation cover, 223.2% increase in built-up areas,
269.1% increase in bare lands and 70% increase in water bodies within the 20 years. Spatio-temporal analysis of the
three imageries revealed that agricultural lands were largely been taken over by urbanization while vegetation had
rapidly given way to bare lands within the 20 years. It was observed that these changes resulted from anthropogenic
activities, environmental factors and climate change. These result in the loss of farmlands, inadequate food supply,
unemployment, inadequate industrial raw materials, reduction in revenue generated, forest depletion, desertification,
wildlife extinction and temperature increase. While it is recommended that reforestation, land reclamation and
irrigation agriculture should be promoted in the area, it is also suggested that further research should focus on the
impact of climate change on land cover change in the area.
Keywords: Dambatta; GIS; Land cover; Land use; Spatio-temporal changes

terchangeably, however, there is a clear difference

1. Introduction between them. Land cover refers to those things that
Most often, land use and land cover are used in- cover part of the earth’s surface like grassland and

*CORRESPONDING AUTHOR:
David Sesugh Aule, Department of Geography, Nigerian Defence Academy, Kaduna, PMB 2109, Nigeria; Email: [email protected]
ARTICLE INFO
Received: 6 May 2023 | Revised: 21 June 2023 | Accepted: 27 June2023 | Published Online: 17 July 2023
DOI: https://doi.org/10.30564/jgr.v6i3.5707
CITATION
Aule, D.S., Jibril, M.S., Idris, A.H., 2023. Spatiotemporal Analysis of Land Use Land Cover Mapping and Change Detection in Dambatta Local
Government Area. Journal of Geographical Research. 6(3): 18-28. DOI: https://doi.org/10.30564/jgr.v6i3.5707
COPYRIGHT
Copyright © 2023 by the author(s). Published by Bilingual Publishing Group. This is an open access article under the Creative Commons Attribu-
tion-NonCommercial 4.0 International (CC BY-NC 4.0) License. (https://creativecommons.org/licenses/by-nc/4.0/).

18
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

forest. Land use on the other hand refers to the man- cision trees [21-24]. It seems evident that when one
ner in which the land is utilized. Land uses include knows what classes are desired and where they occur
activities such as wildlife management, agricultural (at least a sample), supervised classification strate-
practices, urbanization, recreation, etc. [1,2]. A given gies are preferable. Due to its architecture, planting
land use may take place on one, or more than one date, growing period and plant density, each crop has
piece of land and several land uses may occur on the a unique spectral signature [1,25]. This is used during
same piece of land [3]. Changes in land cover may classification, crop identification and discrimination.
occur when one type of land cover is totally replaced In areas where land is scarce, the need to maintain
by another, while changes in land use may result food production with growing demand is achieved
from the modification of land cover types [4-6]. The via technological changes in land use accomplishing
predominant drivers of these scenarios include an- higher returns per area of land [25]. In places of abun-
thropogenic activities causing land subsidence, dance, land conservation is the main strategy used. A
storms and sea level rise [7-10]. better understanding of land use changes is essential
Remote sensing techniques and geographic to assess and predict its effects on ecosystems and
information system (GIS) can be integrated into society [26].
a single tool to perform different tasks including In Dambata LGA, it is revealed that the produc-
detecting land use and land cover changes [11-13]. tivity of the land has declined by about 12% . This
Many researchers have studied the relationship be- is mostly due to land degradation which translates
tween land surface temperature and land use/land to an economic loss of about N18 million [27,28]. It is
cover NDVI using remote sensing and geographic in- the loss of agricultural land with consequent produc-
formation system (GIS). Amongst these, Campbell [14] tivity reduction and economic loss that has necessi-
studied the impact of urbanization on the thermal tated this study. The study employed contemporary
environment of the Bangkok Metropolitan Area. The methods to detect and also analyze the rate of land
study used normalized difference vegetation index use and land cover changes in the area. Thus, this re-
search made use of land cover (LC) data of Dambat-
(NDVI) to extract land use/land cover information
ta LGA as it checked integrally, the spatial and tem-
from remote sensing images of different time periods
poral changes in land use and the rate of land cover
and then analyzed the surface temperature retrieved
change by analyzing the satellite images of 1997,
from the thermal infrared band. The results showed
2007, and 2017 and also checking the vegetation
that urban/built-up areas expanded dramatically
dynamics within the study area. It is on this basis
with equally decreasing agricultural land. More-
that the research work identified the spatio-temporal
over, temperature differences between the urban/
changes in land use and land cover (LULC) that oc-
built-up and the surrounding rural areas significantly
curred in Dambatta LGA between 1997 and 2017.
widened. Similar studies with related results were
conducted [15-20].
The use of remotely sensed data in the form of 2. Materials and methods
satellite imagery as a source of reliable information
2.1 Study area
for surveying, classification, mapping and moni-
toring cannot be over-emphasized. Land use/cover Dambatta Local Government Area is positioned in
mapping is one of the commonest uses of remotely the Northern part of Kano State. The approximate ge-
sensed data. It has been extensively used in super- ographic location of the study area is between latitude
vised image classification where a prior knowledge 12°25’59” N and 12°30’00” N of the equator and
of all cover types to be mapped within the classified longitude 8°30’00” E and 8°50’00” E of the Green-
scene is assumed [1]. Other variants of supervised wich Meridian [27,28] (Figure 1). It has a total land
classification developed in recent years include de- area of 732 km2 and lies within the “wet and dry”

19
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 1. Dambatta LGA showing its localities.

Source: Researchers’ field work.

climate with more dry months than wet months [29,30]. for this study because of their wider swath width
The climate of the study area is categorized under which is required to display a larger area of Dam-
tropical savanna in Koppen’s classification. The batta LGA in a single footprint. Hence, the number
study area falls within the Sudan savannah as well as of footprints required for the study (per year) was
the chad formation underlain by sedimentary rocks reduced to two.
of criterion origin. Agriculture forms the dominant
land use. 2.3 Image classification

A supervised classification procedure was used

2.2 Data collection techniques
in the analysis of data. False color composites were
The coordinates of the control points were ob- developed for the individual images using three out of
tained from a field survey using GPS device (GPSM- the respective bands in accordance with Jibril et al. [27,28].
ap 76 CSx), while the spatial data was obtained from These composites were used to separately classi-
NASRDA, Abuja. These data sets were captured as fy 1997, 2007 and 2017 images. Sample sets were
shown in Table 1. Landsat imageries were chosen created for the individual land cover classes, using

Table 1. Landsat imageries for the study period.

Landsat imageries
Sensor Date Resolution Source Path/Row Satellite
MSS/TM 21st December, 1997 30 m NARSDA 188/050 188/051 Landsat5
ETM+ 4th November, 2007 30 m NARSDA 188/050 188/051 Landsat7
OLI 24th November, 2017 30 m NARSDA 188/050 188/051 Landsat8
Source: Authors’ fieldwork (2018).

20
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

the ‘create’ icon on the sub-set image of the study er (LC) of Dambatta LGA between 1997-2017. Finally,
area. The supervised approach was found useful in descriptive statistical tools were used to analyze the
selecting the calibrated pixels. Training sets were spatio-temporal changes identified.
cross-checked in the field to validate the various land
cover classes for change detection assigned to each 3. Results
training sample set on the image. Thus, a classifica-
tion scheme was developed for this research. The results of the analyses were corroborated by
Subsequently, a maximum likelihood classifier the ground truthing exercise and the satellite images
(MLC) was used in this study because it uses a prob- of 1997, 2007, and 2017. Consequently, some maps
ability density function which enables it to accurate- were achieved and illustrated with Landsat MSS as
ly classify land cover categories that have residual of 1997, Landsat ETM+ as of 2007 and Landsat OLI
ambiguity existing between overlapping classes in as of 2017 respectively (Figure 2). Figure 2 also
the measurement space. This ensured accurate classi- shows the standard color code used to represent the
fication in the analysis. Thus, the applied MLC in IL- five (5) LU LC classes. The green color represents
WIS academy software assumed that the image data agricultural land and the light green represents vege-
exhibits normal distribution and thus, pixels should tation cover. The brown color represents a developed
be made up of a single land use type. MLC allocates area, the milk color represents bare land while blue
pixels in the image [31]. represents the water body.

2.4 Accuracy assessment and field validation 3.1 Temporal analysis of land use land cover
change in Dambatta LGA between 1997 and 2017
The accuracy of classification is a function of the
number of GCPs used for training. A useful rule of Land use (LU) land cover (LC) levels in Dam-
thumb is 30*n (n = number of bands) [32]. The rule was natta LGA in 1997
applied in this study and sampling was done using strat-
The agricultural land as at 1997 covered an esti-
ified random technique on the LC maps. The accuracy
mate of about 51.7% of the lands in the area which
of the classification exercise was measured by ana-
was the highest spatial coverage, followed by vege-
lyzing both commission and omission errors that were
tation cover at about 27.5%, developed area covered
supposed to have occurred during image classification.
about 13.5%, bare land stood at 5.7%; while water
Out of the 700 GCPs, 400 were used for the accuracy
body took 1.3% of the total land area (Table 2). This
assessment. It was assessed for each land cover type
implies that, as at 1997, the level of anthropogenic
separately using an accuracy index (AI) that incorpo-
disturbance in Dambatta LGA was minimal, and
rated both omission and commission error into a single
land degradation was very low with abundant agri-
summary value: AI = ((n-o-c)/n) × 100.
cultural farmlands and vegetation cover in virtually
all corners of the LGA as shown by the statistical
2.5 Data analysis
distribution of LU/LC. This, however, is an estimate
Data processing was conducted using post-classifi- and therefore not very suitable for conclusions, gen-
cation change detection technique in Integrated Land eralizations and decision making. A confusion matrix
and Water Information System ILWIS 5.2, and later was created and used to assess the accuracy of image
converted to shape files which was imported to Arc classification (Table 3).
Map 10.2 GIS software. Landsat images of 30 metres It should be noted that, during classification,
resolution covering the study area in 1997, 2007, and one-pixel value can enter another. Therefore, no im-
2017 were used for land cover classification to identify age classification can be 100% accurate. This is why
the spatio-temporal changes in land use (LU) land cov- Table 4 was created to establish the accuracy or oth-

21
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 2. Land cover changes in Dambatta LGA between 1997-2017.

Source: Authors’ fieldwork (2018).

erwise of the classification. Land use (LU) land cover (LC) change in
The diagonal columns in Table 3 were added to Dambatta LGA between 1997 and 2007
obtain 340. The rows and columns numbers were The land cover distribution as spatially displayed
also summed up to obtain 400. Therefore, the overall in Figure 2 was represented in percentages in Table
accuracy was computed as (340/400 × 100) 85%. 5. The geospatial analysis of the classified Land-
This level of accuracy is indicative of an effective sat ETM + image as at 2007 shows that vegetation
classification result. This method was repeated for cover occupied 25.1% of the total land in the study
2007 and 2017 image classification. area. This was followed by agricultural land at about
Table 2. Estimated LU/LC distribution as of 1997. 24.5%, developed area had increased significantly to
S/No LC observed 1997(km2) % 33.4%, bare land had also increased to 13.7% while
1 Agricultural land 379 51.7 water body covered 3% of the total land mass as at
2 Vegetation cover 202 27.5 2007.
3 Developed area 99 13.5 It was also observed that within a period of 10
4 Bare land 42 5.7 years (1997 to 2007), agricultural land had decreased
5 Water body 10 1.3 significantly by over 50%, while vegetal cover de-
Total 732 100 creased by 10%. Developed area and bare land have
Source: Authors’ fieldwork (2018). increased by 147% and 140% respectively. This

Table 3. Error (confusion) matrix for landsat MSS 1997.

Classified data Referenced data
Agricultural land Vegetation cover Developed area Bare land Water body Total
Agricultural land 182 5 0 1 2 190
Vegetation cover 5 42 1 1 1 50
Developed area 4 10 50 6 0 70
Bare land 9 4 1 31 0 45
Water body 7 3 0 0 35 45
Total 207 64 52 39 38 400
Source: Authors’ fieldwork (2018).

22
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 4. Accuracy assessment for landsat MSS 1997. tion based on observation. Table 6 reveals the quali-
Class name Producer’s accuracy User’s accuracy ty of the digital classification of the landsat imagery
of the study area as at 2007. Producer’s accuracy vis-
Agricultural land 95.7% 87.9%
à-vis user’s accuracy was employed to arrive at the
Vegetation cover 84% 65.6% overall accuracy to ascertain the strength of the clas-
Developed area 68.8% 96.1% sification. Therefore, the overall accuracy of 93.7%
Bare land 71.1% 79.4% was computed. This level of accuracy is indicative of
an effective classification result (Table 7).
Water body 77.7% 92.1%

Overall accuracy 85% Land use (LU) land cover (LC) change in
Dambatta LGA between 2007 and 2017
Source: Authors’ fieldwork (2018).
Supervised classification was also carried out on
Table 5. Land cover distribution as of 2007. the Landsat OLI satellite image of 2017 as shown
in Figure 2. The spatial distribution of land cover
S/NO LC observed 2007 (km2) Percentage (%)
features resulting from the supervised classification
1 Agricultural land 180 24.5
of Landsat OLI satellite image of 2017 is shown in
2 Vegetation cover 184 25.1 Table 8. On the basis of the digital image classifi-
3 Developed area 245 33.4 cation of Dambatta LGA as of 2017, bare land has
4 Bare land 101 13.7 increased progressively over time alongside built-up
5 Water body 22 3.0 areas, while agricultural land and natural vegetation
cover have continuously decreased over time. The
Total 732 100
implication of this result is that anthropogenic ac-
Source: Authors’ fieldwork (2018).
tivities and environmental factors like drought, less
depicts that agricultural land and vegetal cover are rainfall, wind and temperature have increased over
time and have affected vegetal cover in the environ-
giving way for built-up areas and bare lands.
ment leading to an increase in bare land and a conse-
It could be deduced from the observations that
quential effect of land degradation. This has further
built-up areas are increasing obviously as a result led to a reduction in agricultural land. The accuracy
of anthropogenic activities, while bare lands could assessment of the 2017 Landsat OLI satellite image
be increasing as a result of anthropogenic activities computed is 83.25% (See Tables 9 and 10). This
or environmental factors like drought, less rainfall, result is indicative of an accurate and effective clas-
wind and temperature. This, however, is an estima- sification process.
Table 6. Error (confusion) matrix for landsat ETM + 2007.

Classified data Referenced data

Agricultural land Vegetation cover Developed area Bare land Water body Total
Agricultural land 90 8 0 1 1 100
Vegetation cover 5 68 1 1 1 76
Developed area 0 2 20 0 0 22
Bare land 0 2 0 99 1 102
Water body 1 1 0 0 98 100
Total 96 81 21 101 101 400
Source: Authors’ fieldwork (2018).

23
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 7. Accuracy assessment for landsat ETM + 2007. concentrated at the western and eastern parts (Figure
Class name Producer’s accuracy User’s accuracy 2). While bare lands occupied 5.7% of the land area
Agricultural land 90.0% 93.7% in 1997, 13.7% in 2007 and 21.1% in 2017, built up
Vegetation cover 89.4% 83.9% areas had 13.5% of the total land area in 1997, 33.4%
Developed area 90.9% 95.2% in 2007 and 43.7% in 2017 (Table 12). This means
Bare land 97.0% 98.0% that bare lands had positive change differentials of
Water body 98.0% 97.0% 8% between 1997 and 2007, and 7.4% between 2007
Overall accuracy 93.7% and 2017. The built-up areas on the other hand had
Source: Authors’ fieldwork (2018).
positive change differentials of 19.9% between 1997
and 2007 and 10.3% between 2007 and 2017. Gen-
erally, bare lands had increased by 269.1%, built-up
Table 8. Land cover distribution as of 2017.
areas had increased by 223.2%, and water bodies had
S/NO LC observed 2017 (km2) Percentage (93.7%)
increased by 70% within the 20 years. The impli-
1 Agricultural land 130 17.7 cation is that the bare lands and built-up areas were
2 Vegetation cover 110 15.0 increasing continuously throughout the period of the
3 Developed area 320 43.7 study. The increase in both bare lands and developed
4 Bare land 155 21.1 areas is not concentrated in one part of the area. The
5 Water body 17 2.3 proportion covered by bare lands and built-up areas
Total 732 100 increased northwards of the southern part of the
Source: Authors’ fieldwork (2018). area and southwards of the northern part with few
concentrations in the centre of the area (Figure 2).
3.2 Spatial analysis of land use land cover change Very few water bodies were found around the central
in Danbatta LGA between 1997 and 2017 region of the imageries in 1997. A few more chan-
nels of water were found around the western, north
Table 2 shows that Dambatta LGA had 732 km2 western and south eastern regions besides the initial
of land in 1997. It also presents the result of spatial streams of the central region in 2007. The waters of
variation and distribution of land use land cover cat- the south western part of the land, however, disap-
egories within the spatio-temporal frame from 1997 peared in 2017. Thus, the water bodies in the area
to 2017. Table 11 shows that bare land and built-up appear to be increasing and decreasing from 1997 to
areas witnessed remarkable increase which were also 2017.

Table 9. Error (confusion) matrix for landsat OLI 2017.

Classified data Referenced data

Agricultural land Vegetation cover Developed area Bare land Water body Total
Agricultural land 80 20 2 5 3 110
Vegetation cover 10 40 1 1 3 55
Developed area 0 8 89 3 0 100
Bare land 2 8 1 39 0 50
Water body 0 0 0 0 85 85
Total 92 76 93 48 91 400
Source: Authors’ fieldwork (2018).

24
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 10. Accuracy assessment for landsat OLI 2017. to 1997 with just a few scattered bare lands and more
Class name Producer’s accuracy User’s accuracy of vegetation cover with agricultural land (Figure
Agricultural land 72.7% 86.9% 2). Less vegetation and agricultural land were visible
Vegetation cover 72.7% 52.6% in 2007 and the vegetation located in the northern
Developed area 89.0% 95.6% and southern area of the study area had been lost
Bare land 78.0% 78.0% to bare land and built-up area in 2007. In 2017, the
Water body 100% 100% area covered by vegetation decreased in both axis of
Overall accuracy 83.25% the study area when compared to 2007. Generally,
Source: Author’s fieldwork (2018). agricultural lands reduced by 66.8% and vegetation
cover reduced by 45.5% within the 20 years.
Agricultural land and vegetation cover decreased
significantly over the period of study and were scat-
4. Discussion
tered around the study area. Agricultural lands occu-
pied 51.7% of the total land area in 1997, 24.5% in The spatio-temporal analysis reveals a number
2017 and 17.7% in 2017 while vegetation cover had of changes. Initially (1997), the proportions of the
27.5% in 1997, 25% in 2007 and 15% in 2017 (Table land area were used in order of magnitude between
12). The water bodies in the local government area agriculture, vegetation, urbanization, bare lands and
took 1.3% of the total land area in 1997, 3.0% in water bodies. However, the year 2007 had a remark-
2007 and 2.3% in 2017. The comparison of the three able change, bringing the order to developed areas,
satellite imageries (1997, 2007 and 2017) revealed vegetation, agricultural land, bare lands and water
a variation in the pattern of land cover. In 2007, the bodies. It means that development in the area had
proportion covered by bare land increased southward rapidly taken away almost half of the agricultural
of the northern area and northward of southern axis lands between 1997 and 2007. For example, the ag-
with few concentrations at the center when compared ricultural land left for cultivation is not enough to
Table 11. Land cover changes in Dambatta LGA between 1997-2017.

Changed Overall change

1997 2007 2017 Changed area
LULC types area (km2) 1997- (km2)
(km2) (km2) (km2) (km2) 2007-2017
2007 1997-2017
Agricultural land 379 180 130 –199 –50 –249
Vegetation cover 202 184 110 –18 –74 –92
Developed area 99 245 320 +146 +75 +221
Bare land 42 101 155 +59 +54 +113
Water body 10 22 17 +12 –5 +7
Source: Authors’ fieldwork (2018).

Table 12. Percentage of land cover changes in Dambatta LGA between 1997-2017.
LC types 1997 (km2) % 2007 (km2) % 2017 (km2) %
Agricultural land 379 51.7 180 24.5 130 17.7
Vegetation cover 202 27.5 184 25 110 15.0
Developed area 99 13.5 245 33.4 320 43.7
Bare land 42 5.7 101 13.7 155 21.1
Water body 10 1.3 22 3.0 17 2.3
Total 732 100 732 100 732 100
Source: Author’s wield work (2018).

25
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

produce food in quantities that are enough to feed all in almost all parts of Danbatta Local Government
the people in the study area. The implication of this area, especially the northwestern, central and south
result is that most people in the area may partly de- eastern parts.
pend on imported food for survival. This may result The strong and consistent occurrence of the
in famine due to the economic status of the people. changes is most likely due to anthropogenic activities
Furthermore, the reduction in agricultural land has like deforestation, urbanization, and environmental
negative economic implications for the populace. factors like drought, less rainfall, wind, temperature,
A greater number of farmers would have been left erosion and climate change consequently. These
jobless. The government would also record a huge results no doubt have negative implications for
reduction in revenue generated from the sale of farm farmers, ruler dwellers, urban development boards,
produce. Industries that depend on agricultural prod- forest reserves and other policy-makers. This paper,
ucts as raw materials for their products would rather therefore, suggests that further research should be
have to import such raw materials in the midst of conducted focusing on the impact of climate change
the scarcity created by agricultural land inadequacy. on land cover change in the area. It also suggests that
This will lead to a consequent rise in the prices of reforestation; land reclamation and irrigation agricul-
products in the market. ture should be promoted in the area.
In 2017, there were still further changes, bringing
the order to developed areas, bare lands, agriculyural
land, vegetation and water bodies. This time, bare
Author Contributions
lands had taken over the lands previously occupied David Sesugh Aule: Besides serving as the corre-
by vegetation cover. The implication is that the land sponding author, David Sesugh Aule performed the
has been exposed to erosion and its attendant conse- analysis, wrote the abstract and introduction. He also
quences. This explains why there is much land left did much of the discussion ensuring the application
unoccupied in the midst of rapid development. Sand of the results to the study area in various ways.
mining for building activities would have also ex- Mamman Saba Jibril: Mamman Saba Jibril per-
posed the land to erosion leaving a large portion of formed the supervisory role. He guided the ideas that
the land almost useless. In fact, continuous erosion culminated in the production of this paper from the
on the exposed soil surfaces is more likely to dis- beginning to the end. The methodology of this publi-
place settlers on such lands. Loss of vegetation cover cation is his brainchild entirely.
implies loss of timber, wildlife, fire wood, medicine, Ali Husein Idris: Ali Husein Idris worked on data
employment and increased temperature. The loss of collection, processing and spatial analysis. He also
vegetal cover and developed areas are likely to gen- produced the maps that are used in this study.
erate high temperatures which can promote climate
change. This is likely to be the reason why the rivers
Conflict of Interest
and streams in the area continue to increase and de-
crease over the years. The authors declare that they have no conflicts of
interest regarding the publication of this paper.
5. Conclusions
Acknowledgement
The study showed evidence of changes between
1997 and 2017. Spatiotemporal analysis of the three Our profound gratitude and appreciation go
imageries showed that agricultural lands were large- to members of the research team for giving pre-
ly been taken over by urbanization while vegetation cious contributions towards the completion of this
areas had rapidly given way to bare lands. The study research. We also acknowledge members of our
reached the conclusion that these changes are evident various families who supported and encouraged us

26
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

throughout this study. We are most grateful to God, rates of present-day relative sea-level rise in
for giving us guidance and knowledge as well as low-elevation coastal zones: A critical evalua-
good health in constructing this research work. The tion. Ocean Science. 15(1), 61-73.
support of Professor Mamman Saba Jibril who fi- DOI: https://doi.org/10.5194/os-15-61-2019
nanced this work for publication is equally acknowl- [8] Hoegh-Guldberg, O., Masson-Delmotte, V.,
edged. Zhai, P., et al., 2018. Impacts of 1.5°C Glob-
al Warming on Natural and Human Systems
References [Internet]. Available from: http://hdl.handle.
net/10138/311749
[1] Sesha Sai, M.V., Ramana, K.V., Hebbar, R. (edi-
[9] Mentaschi, L., Vousdoukas, M.I., Pekel, J.F.,
tors), 2007. Satellite remote sensing for efficient
et al., 2018. Global long-term observations of
coastal agriculture. International Symposium on
coastal erosion and accretion. Scientific Reports.
Management of Coastal Ecosystems; 2007 Oct
8(1), 12876.
27-30; Kolkata.
DOI: https://doi.org/10.1038/s41598-018-30904-w
[2] Stroosnijder, L., Fresco, L.O., Duiker, S.W.,
[10] Pendergrass, A.G., 2018. What precipitation is
1994. Imaginable futures, a contribution to
extreme?. Science. 360(6393), 1072-1073.
thinking about land use planning. The future of
DOI: https://doi.org/10.1126/science.aat1871
land: Mobilizing and integrating knowledge for
[11] Jadkowski, M.A., Howard, R.R., Brostuen, D.E.,
land use options. John Wiley and Sons: Chich-
1990. Application of SPOT data for regional
ester. pp. 1-8.
growth analysis and local planning. Photogram-
[3] Kebede, G., 2010. GIS Based Surface Irrigation
Potential Assessment of River Catchment for metric Engineering and Remote Sensing. 56(2),
Irrigation Development in Dale Woreda, Sidama 175-180.
Zone, SNNP, School of Natural Resource and [12] Treitz, P.M., Howard, P.J., Gong, P., 1992. Ap-
Environmental Engineering, Haramaya Univer- plication of satellite and GIS technologies for
sity land-cover and land-use mapping at the rural-ur-
[4] Turner, I.I., Ross, R.H., Skole, O.L., 1993. ban fringe: A case study. Journal of Photogram-
Relating Land Use and Global Land-Cover metric Engineering and Remote Sensing. 58,
Change: A Proposal for an IGBP-HDP Core 439-448.
Project [Internet]. A Report from the IGBP/ [13] Dangulla, M., Abd Munaf, L., Mohammad, F.R.,
HDP Working Group on Land-Use/Land-Cov- 2020. Spatio-temporal analysis of land use/land
er Change. Available from: http://ciesin.org/ cover dynamics in Sokoto Metropolis using
docs/008-105/008-105.html multi-temporal satellite data and land change
[5] Lillesand, T., Kiefer, R.W., Chipman, J., 2015. modeller. The Indonesian Journal of Geography.
Remote sensing and image interpretation, 7th 52(3), 306-316.
Edition. John Willey and Sons: New York. pp. [14] Campbell, J.B., Wynne, R.H., 2002. Introduc-
804. tion to remote sensing, 3rd edition. The Guilford
[6] Obiahu, O.H., Yan, Z., Uchenna, U.B., 2021. Press: New York. pp. 621.
Spatiotemporal analysis of land use land cover [15] Star, J.L., Estes, J.E., 1993. Remote Sensing and
changes and built-up expansion projection in GIS Integration: Towards a Prioritized Research
predominantly dystric nitosol of Ebonyi state, Agenda (93-4) [Internet] [cited 2021 Mar 21].
Southeastern, Nigeria. Environmental Challeng- Available from: https://escholarship.org/uc/
es. 4, 100145. item/675609bw
DOI: https://doi.org/10.1016/j.envc.2021.100145 [16] Focardi, S., Loiselle, S.A., Mazzuoli, S., et al.,
[7] Keogh, M.E., Törnqvist, T.E., 2019. Measuring 2008. Satellite-based indices in the analysis of

27
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

land cover for municipalities in the province of Workshop on Satellite Remote Sensing (Nige-
Siena, Italy. Journal of Environmental Manage- riaSat-1) and GIS: A Solution to Sustainable Na-
ment. 86(2), 383-389. tional Development Challenges; 2004 Jun 15-
[17] Cihlar, J., 2000. Land cover mapping of large 17.
areas from satellites: Status and research prior- [26] Dolman, A.J., Moors, E.J., Elbers, J.A., 2002.
ities. International Journal of Remote Sensing. The carbon uptake of a mid latitude pine forest
21(6-7), 1093-1114. growing on sandy soil. Agricultural and Forest
[18] Jansen, L.J., Di Gregorio, A., 2002. Parametric Meteorology. 111(3), 157-170.
land cover and land-use classifications as tools [27] Jibril, M.S., Aule, D.S., Idris, A.H., 2022. Land
for environmental change detection. Agriculture, cover changes in Danbatta local government
Ecosystems & Environment. 91(1-3), 89-100. area of Kano State, Nigeria. AFRIGIST Jour-
[19] Rogan, J., Chen, D., 2004. Remote sensing tech- nal of Land Administration and Environmental
nology for mapping and monitoring land-cover Management. 2(1), 27-38.
and land-use change. Progress in Planning. [28] Jibril, M.S., Aule, D.S., Idris, A.H., 2022. As-
61(4), 301-325.
sessment of land degradation in Dambatta local
[20] Weng, Q., 2009. Thermal infrared remote sens-
government area using remote sensing tech-
ing for urban climate and environmental studies:
niques. Advances in Remote Sensing. 11(4),
Methods, applications, and trends. ISPRS Jour-
167-181.
nal of Photogrammetry and Remote Sensing.
DOI: https://doi.org/10.4236/ars.2022.114010
64(4), 335-344.
[29] Mohammed, M. A., 2015. Impact of Soil Micro-
[21] Hansen, M.C., DeFries, R.S., Townshend, J.R.,
bial Respiration on Atmospheric Carbon Under
et al., 2000. Global land cover classification at 1
Different Land Use in Part of Kano State, North-
km spatial resolution using a classification tree
ern Nigeria. International Journal of Animal Bi-
approach. International Journal of Remote Sens-
ology. Vol. 1, No. 5, 2015, pp. 158-164.
ing. 21(6-7), 1331-1364.
[22] Foody, G.M., 1998. Sharpening fuzzy classi- [30] Olofin, E.A., 1987. Some effects of Tiga Dam
fication output to refine the representation of on the environmental downstream in Kano Riv-
sub-pixel land cover distribution. International er Basin [Ph.D. thesis]. Zaria: Department of
Journal of Remote Sensing. 19(13), 2593-2599. Geography, A.B.U.
[23] Mannan, B., Roy, J., Ray, A.K., 1998. Fuzzy ART- [31] Smith, D.R., Chandra, S., Barth, R.F., et al.,
MAP supervised classification of multi-spectral 2001. Quantitative imaging and microlocaliza-
remotely-sensed images. International Journal of tion of boron-10 in brain tumors and infiltrating
Remote Sensing. 19(4), 767-774. tumor cells by SIMS ion microscopy: Relevance
[24] Van Der Meer, F., 1995. Spectral unmixing of to neutron capture therapy. Cancer Research.
landsat thematic mapper data. International 61(22), 8179-8187.
Journal of Remote Sensing. 16(16), 3189-3194. [32] Jensen, J.R., 1996. Introductory digital image
[25] Raji, B.A., 2004. Agricultural land use planning processing: A remote sensing perspective, 2nd
and management in Kadawa irrigation scheme, Edition. Prentice Hall, Inc.: Upper Saddle River,
Kano State. Paper presented at the National NJ.

28
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Journal of Geographical Research

https://journals.bilpubgroup.com/index.php/jgr

ARTICLE

Application of Vegetation Indices for Detection and Monitoring Oil Spills

in Ahoada West Local Government Area of Rivers State, Nigeria
Jonathan Lisa Erebi* , Egirani E. Davidson

Department of Geology, Niger Delta University, Wilberforce Island, Bayelsa State, 560103, Nigeria

ABSTRACT
The study evaluated the environmental effects of an oil spill in Joinkrama 4 and Akimima Ahoada West LGA,
Rivers State, Nigeria, using various vegetation indices. Location data for the spill were obtained from the Nigeria Oil
Spill Detection and Response Agency, and Landsat imagery was acquired from the United States Geological Survey.
Three soil samples were collected from the affected area, and their analysis included measuring total petroleum
hydrocarbons (TPH), total hydrocarbons (THC), and polycyclic aromatic hydrocarbons (PAH). The obtained data were
processed with ArcGIS software, utilizing different vegetation indices such as the Normalized Difference Vegetation
Index (NDVI), Atmospheric Resistant Vegetation Index (ARVI), Soil Adjusted Vegetation Index (SAVI), Green
Short Wave Infrared (GSWIR), and Green Near Infrared (GNIR). Statistical analysis was performed using SPSS and
Microsoft Excel. The results consistently indicated a negative impact on the environment resulting from the oil spill.
A comparison of spectral reflectance values between the oil spill site and the non-oil spill site showed lower values
at the oil spill site across all vegetation indices (NDVI 0.0665-0.2622, ARVI –0.0495-0.1268, SAVI 0.0333-0.1311,
GSWIR –0.183-0.0517, GNIR –0.0104-–0.1980), indicating damage to vegetation. Additionally, the study examined
the correlation between vegetation indices and environmental parameters associated with the oil spill, revealing
significant relationships with TPH, THC, and PAH. A t-test with a significance level of p < 0.05 indicated significantly
higher vegetation index values at the non-oil spill site compared to the oil spill site, suggesting a potential disparity
in vegetation health between the two areas. Hence, this study emphasizes the harmful effect of oil spills on vegetation
and highlights the importance of utilizing vegetation indices and spectral reflectance analysis to detect and monitor the
impact of oil spills on vegetation.
Keywords: Vegetation indices; TPH; PAH; THC; Oil spill; Impact; Rivers State; Nigeria

*CORRESPONDING AUTHOR:
Jonathan Lisa Erebi, Department of Geology, Niger Delta University, Wilberforce Island, Bayelsa State, 560103, Nigeria; Email: jonathanli-
[email protected]
ARTICLE INFO
Received: 28 June 2023 | Revised: 27 July 2023 | Accepted: 1 August 2023 | Published Online: 7 August 2023
DOI: https://doi.org/10.30564/jgr.v6i3.5817
CITATION
Erebi, J.L., Davidson, E.E., 2023. Application of Vegetation Indices for Detection and Monitoring Oil Spills in Ahoada West Local Government
Area of Rivers State, Nigeria. Journal of Geographical Research. 6(3): 29-41. DOI: https://doi.org/10.30564/jgr.v6i3.5817
COPYRIGHT
Copyright © 2023 by the author(s). Published by Bilingual Publishing Group. This is an open access article under the Creative Commons Attribu-
tion-NonCommercial 4.0 International (CC BY-NC 4.0) License. (https://creativecommons.org/licenses/by-nc/4.0/).

29
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

1. Introduction multinational oil companies operating in the region.

Technological advancements, particularly vegetation
Oil spills in the Niger Delta, particularly in spectral indices and remote sensing, offer valuable
Joinkrama 4 and Akinima, have emerged as a criti- tools for detecting and monitoring the impacts of oil
cal global issue with far-reaching consequences for
spills in these environments [7]. Remote sensing tech-
ecosystems, wildlife, and human livelihoods [1] The
niques, including vegetation spectral indices, have
discovery of crude oil in Oloibiri in 1956 marked the
proven effective in detecting and monitoring the
beginning of a series of oil spill incidents, both inten-
effects of hydrocarbon leaks on vegetation health [8].
tional and unintentional, which have led to the devas-
Changes in leaves, stems, and trunks serve as indica-
tation of rivers, seas, and marshes, posing significant
tors of plant responses to oil contamination or stress [9].
environmental hazards [2]. The Niger Delta, acknowl-
Continuous monitoring of pipelines and oil leaks at
edged as one of the world’s ten most important wet-
oil production sites is crucial to mitigate oil contami-
lands and marine ecosystems, has suffered extensive
nation in rivers, seas, oceans, and marshes.
environmental ramifications due to the presence of
numerous oil industries [3]. While these industries
have contributed to the growth and development 2. Material and methods
of the region, they have also resulted in adverse
2.1 Location of the study area
impacts. The relentless activities of restless youths
who vandalize pipelines and the rising number of The research focuses on the settlements of Joink-
illegal refineries in oil-producing areas have sub- rama 4 and Akinima in Rivers State, Nigeria’s Ahoa-
stantially amplified environmental contamination [1]. da West Local Government Area. These towns are
The sensitivity of vegetation to hydrocarbons varies located on the eastern bank of the Orashi River, a
depending on the type and quantity of chemicals significant Niger Delta river (Figure 1). Joinkrama
involved, as well as the type of vegetation [4]. The 4 is located at roughly 4.8265° N, 6.0665° E, where-
impact of oil spills poses a significant threat to the as Akinima is located at approximately 5.1046° N,
environment in Nigeria, especially in the Niger Del- 6.4529° E. The area’s geographical location is large-
ta. If not effectively managed, it could lead to the ly flat, with heights varying from below sea level in
complete destruction of ecosystems in the region the southwest to roughly 39 meters inland [10]. The
where oil spills have become alarmingly frequent [5]. region is easily accessible by road and footpath, and
Communities in this area suffer from the degrading it has multiple hydrocarbon flow stations operated
effects of oil spills, as the once-vibrant mangroves by Shell Petroleum Development Company (SPDC)
that provided fuelwood and habitat for biodiversity and Nigeria Agip Oil Company (NAOC) [10]. The
have been depleted and rendered lifeless due to oil region is drained by various tributaries and creeks,
toxicity. Marine life has also been adversely affected including Kolo Creek and Taylor Creek, which are
by oil spills, resulting in contaminated seafood and connected to the Orashi River.
subsequent health risks associated with human con-
sumption. The oil leak has had a significant influence 2.2 The geology of the area
on the health, social, and economic aspects of the
affected communities, emphasizing the importance The research area is geologically located on the
of effective steps to avoid and handle oil disasters south-western side of the Niger Delta region, which
in Nigeria’s Niger Delta region [6]. Farmlands and was produced by a failed rift junction between the
sources of potable water are destroyed, further im- South American and African plates. The rifting
pacting coastal fishing activities. The recurrent in- occurred between the late Jurassic and mid-Cre-
cidents of oil spills in the Niger Delta have created taceous periods [11]. The Niger Delta Basin has an
significant tension between the local population and area of more than 105,000 km2 [12]. The Akata For-

30
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

mation, Agbada Formation, and Benin Formation 2.3 Data collection for oil spill
are among the geological formations studied in the
The satellite data were collected from the United
area. According to Short and Stauble [13], the Akata
States Geological Survey (USGS). The data includes
Formation, which ranges in age from the Paleocene
Landsat-5 acquired on September 12, 1990, with a
to the Holocene, is the Niger Delta region’s base
spatial resolution of 30 meters, and Landsat-8 ac-
lithostratigraphic unit. It is made up of thick shales,
quired on December 12, 2021, with a similar spatial
turbidite sands, and traces of silt and clay. Near the
resolution. The source for both datasets is the USGS
interface with the underlying Agbada Formation,
Earth Explorer website. Additionally, oil spill records
the formation comprises high-pressure, low-density
and corresponding GPS locations from November
deep marine deposits containing plant remains. It
2020 to April 2021 were obtained from the Nigeria
has a varied microfauna, with planktonic foraminif-
Oil Spill Detection and Response Agency (NOS-
era accounting for a sizable percentage, indicating a
DRA) archives, available at https://oilspillmonitor.
shallow maritime shelf depositional environment [14].
ng/#/.
The sand and silt streaks inside the formation pro-
vide evidence of high-energy deposition as the delta
2.4 Collection of data for soil
moved into the sea. The Akata Formation extends in
thickness from 0 to 6,000 meters and is mostly un- A total number of three (3) soil samples were col-
derwater, not visible at the seashore [15]. lected with the aid of a hand auger two were collect-

Figure 1. Study area map.

31
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

ed in Joinkrama 4, one in Akinima and Soil samples employed in the study, utilizing the spectral bands of
were collected from the outer surface and were ana- Landsat-5 and Landsat-8. For Landsat-8, the spectral
lyzed in the laboratory for Polycyclic Aromatic Hy- bands used for vegetation indices are as follows: vis-
drocarbon, (PAHs), Total Petroleum Hydrocarbons ible blue (0.450 to 0.515 µm), visible green (0.525 to
(TPH) and Total Hydrocarbon Content (THC). 0.6 µm), visible red (0.630 to 0.68 µm), near-infra-
red (0.845 to 0.885 µm), and short-wave infrared I
2.5 Limitation (1.56 to 1.66 µm), all with a resolution of 30 meters.
In the case of Landsat-5, the vegetation indices are
In order to minimize the effect of cloud cover, in derived from the following spectral bands: visible
Landsat-8, this study is limited to the month of De- blue (0.450 to 0.515 µm), visible green (0.525 to 0.6
cember 2021 as there is less cloud cover at this time µm), visible red (0.630 to 0.69 µm), near-infrared
of the year. (0.77 to 0.90 µm), and short-wave infrared 1 (1.55
to 1.75 µm), also with a resolution of 30 meters.
2.6 Data analysis/processing By analyzing these vegetation indices derived from
Landsat imagery, researchers can assess the extent of
PAH chemicals were analyzed using gas chro-
oil spills and their impact on the environment, pro-
matography, TPH and phenol were measured using
viding valuable insights for environmental studies.
a photometer, and THC content was determined
Atmospheric correction refers to the process
through spectrophotometer readings. These analyti-
of adjusting the recorded satellite imagery to ac-
cal techniques and instruments allowed for the accu-
count for atmospheric effects and obtain accurate
rate identification and quantification of these hydro-
measurements of the Earth’s surface. The equations
carbon compounds in various samples.
for atmospheric correction of Landsat-8 imager
The oil spill data was in degrees, minutes and
from Digital Number (DN) to Top of Atmosphere
seconds and was imported into Microsoft Excel and
(TOA) reflectance using reflectance rescaling coeffi-
converted to degree decimal before it was loaded
cients from Landsat-8 metadata file are provided in
into the Geographical Information System envi-
Equations (1) and (2) [16-18].
ronment in Database Format to produce a sample
Equation (1): TOA reflectance without correction
location map using Arc GIS software. The Arc GIS for solar angle (ρλ’) can be calculated using the for-
spatial analyst extension was used to generate the mula:
thematic maps using Map Algebra in a raster calcu-
pλ = MpQcal + Ap (1)
lator, and various vegetation indexes such as NDVI,
Here, pλ represents TOA planetary reflectance,
SAVI, ARVI2, G-NIR and G-SWIR were derived.
Mp is the band-specific multiplicative rescaling fac-
Finally, SPSS (Statistical Package for the Social Sci-
tor, Aρ is the band-specific additive rescaling factor,
ence) and IBM were used for statistical analysis such
and Qcal refers to the quantized and calibrated stand-
as Pearson correlation coefficient and T-test were
ard product pixel values (DN).
done and Microsoft Excel was also used to plot the
Equation (2): TOA reflectance corrected for the
bar.
sun angle can be calculated as follows:
pλ = (pλ) / cos(θsz) = (pλ) / cos(θse) (2)
2.7 Vegetation indices
In Equation (2), θsz represents the solar zenith
The use of Landsat imagery, specifically Land- angle, while θse represents the solar elevation angle.
sat-5 and Landsat-8, is crucial for detecting oil The Normalized Difference Vegetation Index
spills and understanding their environmental impact (NDVI) is a commonly used spectral index that ex-
through vegetation indices derived from broadband hibits distinct characteristics for detecting vegetation
multispectral data. Five vegetation stress factors are stress. It utilizes the red band in the visible spectrum,

32
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

which is sensitive to chlorophyll content, and the In Equation (7), RSWIR represents the reflec-
near-infrared (NIR) band, which indicates healthy tance in the short-wave infrared (SWIR) band.
vegetation conditions. Equation (3) represents the These equations and indices play crucial roles in
calculation for NDVI [19,20]. analyzing satellite imagery and extracting valuable
NDVI = (RNIR – RRED) / (RNIR + RRED) (3) information about vegetation and environmental
Here, RNIR represents the reflectance in the NIR conditions.
band, and RRED represents the reflectance in the red
band. 3. Results and discussion
The Soil Adjusted Vegetation Index (SAVI) was
developed by Huete [21] and Huete [22] to address
3.1 Assessment of oil spill’s impact on the envi-
noise present in NDVI due to canopy background
ronment
and atmospheric conditions. SAVI incorporates an The assessment of an oil spill’s impact Joinkrama
adjustment factor and is useful for correcting soil 4 and Akinima utilized the NDVI. Tables 1 and 2
brightness and atmospheric effects [23]; presented a comparison of NDVI spectral reflectance
SAVI = ((RNIR – RRED) / (RNIR + RRED + L)) values between the oil spill site and a non-oil spill
× (1 + L) (4) site. The results in Figure 2 indicated the presence
In Equation (4), L represents the adjustment fac- of unhealthy plants in the spill-affected area. The re-
tor, typically set to 0.5. sults revealed that the NDVI spectral reflectance val-
The Atmospheric Resistant Vegetation Index 2 ues at the oil spill site were lower, indicating damage
(ARVI2) is designed to be resistant to atmospheric to the vegetation. Figure 3 depicted the spectral
effects and sensitive to a wide range of chlorophyll reflectance, further emphasizing the significant ex-
concentrations. It is influenced by vegetation fraction tent of damage caused by the spill. The Oil Spill Site
and the rate of absorption of photosynthetic solar (OSS) exhibited a range of 0.0665 to 0.2622, while
radiation. Equation (5) represents the calculation for the Non-Oil Spill Site (NOSS) had a value of 0.4522.
ARVI2 [24,25]: These findings demonstrate that the oil spill had a
substantial adverse impact on the studied area’s veg-
ARVI2 = –0.18 + 1.17 × (RNIR – RRED / RNIR
+ RRED) (5) etation, as evidenced by the lower NDVI values and
the presence of unhealthy plant growth.
The Green-Near Infrared (G-NIR) index com-
The resulting SAVI values are then compared
bines the green and near-infrared reflectance values
with historical data or a control area to identify re-
to assess plant vigor and characterize vegetation
gions where vegetation cover has been affected by
structure. It has shown potential for discriminating
the oil spill. SAVI proves effective in detecting oil
between vegetation affected by oil spills and unaf-
spills because it provides a quantitative measure of
fected sites [26,27]:
the extent of the damage caused. The provided infor-
G-NIR = (RGREEN – RNIR) / (RGREEN + RNIR) (6) mation indicates a negative impact on vegetation in
In Equation (6), RGREEN represents the reflec- the affected area due to the oil spill. Comparisons of
tance in the green band. SAVI values between the oil spill site and a non-oil
The Green-Short-Wave Infrared (G-SWIR) index spill site, as shown in Tables 1 and 2, further sup-
has the ability to predict and sense nitrogen levels in port the evidence of spill-induced damage in Figure
plants [28] and can discriminate the moisture content 4. Spectral reflectance data in Figure 5 illustrates
of soil and vegetation. It may be useful for detecting that the oil spill site exhibits a lower SAVI range of
changes in vegetation affected by oil spills [29]: 0.0333 to 0.1311, while the non-oil spill site displays
G-SWIR = (RGREEN – RSWIR) / (RGREEN a higher value of 0.2261. This disparity suggests
+ RSWIR) (7) that vegetation in the oil spill site is less healthy

33
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

compared to the non-oil spill site. Therefore, the A comprehensive investigation of the influence of
presented information strongly indicates a significant oil spills on vegetation stress through spectral reflec-
negative impact of the oil spill on the vegetation in tance in GNIR bands was conducted, as evident from
the affected area. the findings presented in Tables 1 and 2 and Figure
The study examined the health of vegetation in 8. These data depict the pre- and post-oil spill GNIR
two distinct areas, one affected by an oil spill and the results, illustrating that the spills have induced stress
other untouched, using the Atmospheric Resistant and harm to the vegetation, resulting in changes in
Vegetation Index (ARVI). The analysis, presented in reflectance in both the green and NIR spectral bands.
Tables 1 and 2, along with Figure 6, indicated the Moreover, the comparison of spectral reflectance
condition of vegetation in these areas. The color-cod- between the oil spill site (OSS) and the non-oil spill
ed images demonstrated that the regions affected by site (NOSS) in Figure 9, provides further insights.
the oil spill exhibited unhealthy vegetation, while The results demonstrate that reflectance values in
the non-oil spill area showed moderate vegetation. the OSS range from –0.0479 to –0.1980, while the
Tables 1 and 2 were utilized to compare the ARVI NOSS exhibits a reflectance of –0.3823. This com-
values between the oil spill site and the non-oil spill parison signifies that the presence of oil spills within
site. The results derived from these tables were then the area significantly impacts the health of vegeta-
used to generate Figure 7, which depicted the spec- tion, leading to detectable changes in spectral reflec-
tral reflectance of the oil spill site (OSS) and the tance that can be closely monitored. Consequently,
non-oil spill site (NOSS). The range of the OSS val- the study underscores the significance of employing
ues, ranging from –0.1022 to 0.1268, indicated the GNIR spectral reflectance as a means to identify and
extent of vegetation damage caused by the oil spill. monitor the effects of oil spills on vegetation health.
In comparison, the NOSS value of 0.3492 provided This knowledge can aid in the development of effec-
a reference point for evaluating the impact of the oil tive strategies to mitigate the damage caused by such
spill on vegetation in the area. incidents.

Figure 2. Vegetation analysis of normalized difference vegetation index (NDVI) 1990 and 2021.

Figure 3. Normalized difference vegetation index for OSS and NOSS 2021.

34
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 4. Vegetation analysis of soil adjusted vegetation index (SAVI) 1990 Figure 5. Vegetation analysis of soil adjusted vegeta-
and 2021. tion index (SAVI) for OSS and NOSS 2021.

Figure 6. Vegetation analysis of atmospheric resistant vegetation index Figure 7. Vegetation analysis of atmospheric resistant
(ARVI) 1990 and 2021. vegetation index (ARVI) for OSS and NOSS 2021.

Figure 9. Vegetation analysis of green-near infrared

Figure 8. Vegetation analysis of green-near infrared (GNIR) 1990 and 2021. (GNIR) for OSS and NOSS 2021.

Figure 11. Vegetation analysis of green short wave

Figure 10. Vegetation analysis of green short wave infrared (GSWIR) 1990 infrared (GSWIR) for OSS and NOSS 2021.
and 2021.

35
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

The findings presented in Tables 1 and 2 demon- correlation coefficient analysis, illustrating the re-
strate the primary objective of the study, which was lationship between various vegetation indices prior
to evaluate how oil spills influence the well-being of to and during an oil spill occurrence. The Pearson
vegetation by analyzing the spectral reflectance with- correlation coefficient assesses the linear connection
in the Green Short Wave Infrared (GSWIR) range. between two variables and ranges from –1 (indicat-
Figure 10 reveals changes in GSWIR reflectance ing a perfect negative correlation) to +1 (indicating a
before and after the oil spill, indicating that the pres- perfect positive correlation), with 0 representing no
ence of the spill causes stress and harm to the vegeta- correlation. Based on the information provided, the
tion, as evidenced by the altered reflectance. Tables Normalized Difference Vegetation Index (NDVI),
4 and 5 compare the GSWIR reflectance between the Atmospheric Resistant Vegetation Index (ARVI),
oil spill site (OSS) and the non-oil spill site (NOSS). and Soil Adjusted Vegetation Index (SAVI) exhibit
Figure 11 illustrates that OSS 1 to OSS3 exhibit a a strong positive correlation of r = 0.9229 before
range of values from 0.05717 to –0.1803, with OSS3 and during the oil spill event. This implies that these
having a higher value than NOSS (–0.3823). This indices are highly associated and move in the same
suggests that the presence of an oil spill has adverse- direction, suggesting their utility in monitoring veg-
ly affected the health of the vegetation in the area, etation changes during and after an oil spill. The
resulting in changes in spectral reflectance in the statement further notes that the Green Short Wave
Short Wave Infrared (SWIR) range. Thus, the study Infrared (GSWIR) index demonstrates a strong pos-
employed GSWIR spectral reflectance as a means to itive correlation of r = 0.7559 before and during the
detect and monitor the impact of oil spills on vegeta-
oil spill event. Although this correlation is not as
tion health. The results underscore the necessity for
robust as the aforementioned indices, it still signifies
effective measures to prevent oil spills and minimize
a positive relationship between the variables. Last-
their detrimental consequences on the environment.
ly, the Green Near Infrared (GNIR) index reveals
Table 1. Vegetation indices spectral reflectance during the spill a strong positive correlation of r = 0.93649 before
in soil sample 2021.
and during the oil spill event. This indicates a high
NDVI GSWIR GNIR correlation between GNIR and the other vegetation
CODE ARVI 2021 SAVI 2021
2021 2021 2021
indices, making it suitable for monitoring vegetation
OSS1 0.1115 –0.0495 0.0558 0.0517 –0.0479
changes throughout an oil spill event. Consequently,
OSS2 0.0665 –0.1022 0.0333 0.0214 –0.0104
these correlations suggest that by combining differ-
OSS3 0.2622 0.1268 0.1311 –0.1803 –0.1980
ent vegetation indices, one can effectively monitor
Note: Oil spill site = OSS.
vegetation changes before, during, and after an oil
Table 2. Vegetation indices spectral reflectance in non-oil spill
spill occurrence.
site 2021. Table 3. Pearson correlation coefficient between vegetation
Vegetation indices NOSS indices before the spill and during the spill.

NDVI 0.4523 Vegetation indices before spill

r Strength Direction
ARVI 0.3492 (1990) and during spill (2021)

SAVI 0.2261 NDVI 0.9229 Strong positive

GSWIR –0.1332 ARVI 0.9229 Strong positive

GNIR –0.3823 SAVI 0.9229 Strong positive
Note: Non-oil spill site = NOSS. GSWIR 0.7559 Strong positive
GNIR 0.9364 Strong positive
3.2 Pearson correlation coefficient analysis
The results from Tables 4-7 indicate that there
Table 3 presents the findings of the Pearson is a significant correlation between the Normalized

36
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Difference Vegetation Index (NDVI) and the envi- Table 7. Pearson correlation coefficient between vegetation
ronmental parameters TPH, THC, and PAH in the indices and PAH (topsoil).

affected oil spill site. Specifically, the results show 2021 r Strength Direction
a strong negative correlation (r = –0.7065) between NDVI vs PAH 0.5827 Strong Positive
NDVI and TPH, a strong positive correlation (r = ARVI vs PAH 0.5827 Strong Positive
0.6968) between NDVI and THC, and a strong pos- SAVI vs PAH 0.5827 Strong Positive

itive correlation (r = 0.5827) between NDVI and GSWIR vs PAH –0.2765 Weak Negative

PAH. These correlations suggest that as the levels of GNIR vs PAH –0.5570 Strong Negative

TPH increase, the NDVI decreases, and as the levels

of THC and PAH increase, the NDVI increases. This 3.3 T-test result for vegetation indices be-
could indicate that higher levels of TPH negatively tween the Oil Spill Site (OSS) and Non-Oil
impact vegetation, while higher levels of THC and Spill Site (NOSS)
PAH positively impact vegetation in this particular
oil spill site. However, it is important to note that A dependent t-test was performed to compare the
correlation does not necessarily imply causation, and Normalized Difference Vegetation Index (NDVI)
further research is needed to understand the specific spectral reflectance between the Oil Spill Site (OSS)
mechanisms behind these relationships. Nonetheless, and Non-Oil Spill Site (NOSS), as indicated in Ta-
these findings could have important implications for ble 8. The results demonstrated a statistically sig-
understanding and mitigating the impacts of oil spills nificant difference between the two sites, with the
on vegetation in affected areas. NOSS mean surpassing the OSS mean. The t-test
was conducted with a significance level (alpha) of
Table 4. Hydrocarbon results in soil.
p < 0.05, and the assumption of normally distributed
Sample code TPH THC PAH
difference scores was confirmed. The mean NDVI
Soil 1 (topsoil) 1.693 4.8 0.045 spectral reflectance for OSS was M = 0.1467 with a
Soil 2 (topsoil) 1.736 4.1 0.03 standard deviation of SD = 0.10250, while the mean
Soil 3(topsoil) 1.691 4.82 0.043 for NOSS was M = 4.523 with a standard devia-
tion of SD = 0.0000. The effect size was considered
Table 5. Pearson correlation coefficient between vegetation
indices and TPH (topsoil). large, with d = 2.98, indicating a strong practical sig-
nificance. Therefore, the findings imply that the pres-
2021 r Strength Direction
ence of an oil spill negatively impacted vegetation
NDVI vs TPH –0.7065 Strong Negative
health, resulting in lower NDVI spectral reflectance
ARVI vs TPH –0.7065 Strong Negative
at the Oil Spill Site (OSS) compared to the Non-
SAVI vs TPH –0.7064 Strong Negative
Oil Spill Site (NOSS). These results underscore the
GSWIR vs TPH 0.4283 Weak Positive necessity for effective measures to prevent oil spills
GNIR vs TPH 0.6840 Strong Positive and minimize their environmental impact.
Table 6. Pearson correlation coefficient between vegetation
In Table 9, the results of a dependent t-test inves-
indices and THC (topsoil). tigating the differences in the Atmospheric Resistant
Vegetation Index (ARVI) between the Oil Spill Site
2021 r Strength Direction
(OSS) and Non-Oil Spill Site (NOSS) were pre-
NDVI vs THC 0.6958 Strong Positive
sented. The research hypothesis aimed to determine
ARVI vs THC 0.6958 Strong Positive
whether the means of ARVI in OSS and NOSS were
SAVI vs THC 0.6958 Strong Positive
equal. The findings indicated a statistically signif-
GSWIR vs THC –0.4147 Weak Negative
icant difference between the mean ARVI scores of
GNIR vs THC –0.6730 Strong Negative OSS and NOSS, with a t-value of 5.163 and a p-val-

37
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

ue less than 0.001. This suggests that the disparity was examined and satisfied prior to conducting the
between the means is unlikely to occur by chance. t-test. This assumption is critical for the validity
Furthermore, the effect size (d) was calculated as of the t-test results. Additionally, an alpha level of
2.98, demonstrating a substantial degree of practical p < 0.05 was employed, which is a common level of
significance. Thus, the results indicate that the mean significance in statistical testing. Overall, the results
ARVI for NOSS is significantly higher than the mean suggest that the GSWIR values were significantly
for OSS. It is important to note that the assumption higher at NOSS compared to OSS, suggesting a po-
of normally distributed difference scores was exam- tential disparity in the presence of oil or other con-
ined and met prior to conducting the analysis. taminants between the two sites. However, the effect
The data in Table 10 show the results of a de- size was weak, implying that the practical signifi-
pendent t-test conducted to compare the Soil Ad- cance of the difference might not be substantial.
justed Vegetation Index (SAVI) spectral reflectance The results depicted in Table 12 demonstrate the
at an Oil Spill Site (OSS) and a Non-Oil Spill Site use of a dependent t-test to examine the differences
(NOSS). The mean SAVI for OSS was 0.0734 with a in Green Near Infrared (GNIR) spectral reflectance
standard deviation of 0.05122, while the mean SAVI between an Oil Spill Site (OSS) and a Non-Oil Spill
for NOSS was 0.02261 with a standard deviation Site (NOSS). The findings indicated a statistically
of 0.0000. The results of the dependent t-test indicat- significant difference between OSS and NOSS, with
ed a statistically significant difference between the a t-value of –5.180 and a p-value less than 0.001.
SAVI values for the OSS and NOSS (t (3) = 5.164, This implies that the NOSS mean was significantly
p < 0.001). The effect size was reported as large (d = higher than the OSS mean for GNIR. The effect size
2.98), signifying a strong practical significance. Pri- was also computed, yielding a large effect size of d =
or to conducting the t-test, it was confirmed that the 3.0. This indicates a strong degree of practical signif-
assumption of normally distributed difference scores icance between the two groups. Prior to conducting
was satisfied. This assumption is crucial for the the analysis, the assumption of normally distributed
difference scores was examined, and the skew and
validity of the t-test results. Furthermore, an alpha
kurtosis levels were found to be –1.46 and 0.00, re-
level of p < 0.05 was employed, which is a common
spectively. These values were within the acceptable
significance level in statistical testing. Consequently,
range for a t-test (i.e., skew < |2.0| and kurtosis <
the results suggest that the SAVI values were signifi-
|9.0|; posten 1984), confirming the satisfaction of the
cantly higher at NOSS compared to OSS, indicating
assumption of normally distributed difference scores.
a potential discrepancy in vegetation health between
Therefore, the results suggest a significant differ-
the two sites.
ence in GNIR spectral reflectance between OSS and
Table 11 presents the outcome of a dependent
NOSS, with NOSS exhibiting a higher mean than
t-test comparing the Green Short Wave Infrared
OSS.
(GSWIR) spectral reflectance at an Oil Spill Site
(OSS) and a Non-Oil Spill Site (NOSS). The mean
GSWIR for OSS was –0.0347 with a standard de- 4. Conclusions
viation of 0.07281, whereas the mean GSWIR for The assessment of an oil spill’s impact on veg-
NOSS was –0.1332 with a standard deviation of etation in a specific area, utilizing the Normalized
0.0000. The dependent t-test revealed a statistically Difference Vegetation Index (NDVI), Atmospheric
significant difference between the SWIR values for Resistant Vegetation Index (ARVI), Soil Adjusted
OSS and NOSS (t (3) = –1.34, p < 0.001). The effect Vegetation Index (SAVI), Green Short Wave Infrared
size was reported as large (d = 0.77), indicating a (GSWIR), and Green Near Infrared (GNIR) spectral
weak practical significance. It is noteworthy that the reflectance, revealed a significant negative impact
assumption of normally distributed difference scores on the health of vegetation due to the oil spill. The

38
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 8. Paired samples statistics and Test for NDVI.

Mean N Std. deviation Std. error mean Skew Kurtosis t

NOSS 0.4523 3 0 0
1.36 0.0000 5.164
OSS 0.1467 3 0.1025 0.05918

Table 9. Paired samples statistics and tests for ARVI.

Mean N Std. deviation Std. error mean Skew Kurtosis t

NOSS 0.3492 3 0.00000 0.00000
1.36 0.0000 5.163
OSS –0.0083 3 0.11993 0.06924

Table 10. Paired samples statistics and tests for SAVI.

Mean N Std. error mean Skew Kurtosis t

NOSS 0.2261 3 0.00000
1.36 0.0000 5.164
OSS 0.0734 3 0.02957

Table 11. Paired samples statistics and tests for SWIR.

Mean N Std. deviation Std. error mean Skew Kurtosis t
NOSS –0.1332 3 0.00000 0.00000
–1.62 0.0000 –1.339
OSS –0.0357 3 0.12611 0.07281

Table 12. Paired samples statistics and tests for GNIR.

Mean N Std. deviation Std. error mean Skew Kurtosis t
NOSS –0.3823 3 0.00000 0.00000
–1.46 0.0000 –5.180
OSS –0.0854 3 0.09927 0.05731

comparisons between the oil spill site (OSS) and tween vegetation indices and environmental param-
non-oil spill site (NOSS) consistently demonstrated eters provide insights into the relationship between
lower vegetation health and damage in the OSS. The oil spill characteristics and vegetation health. These
color-coded representations and spectral reflectance findings highlight the importance of monitoring and
data provided visual evidence of unhealthy and mitigating the impacts of oil spills on vegetation. By
stressed vegetation in the spill-affected area. The utilizing remote sensing techniques and spectral re-
results from various statistical analyses, including flectance analysis, it becomes possible to detect and
dependent t-tests and Pearson correlation coeffi- monitor the extent of damage caused by oil spills.
cients, further supported the findings by showing This knowledge can aid in the development of ef-
significant differences and relationships between fective strategies for preventing and minimizing the
vegetation indices and environmental parameters. detrimental consequences of oil spills on the envi-
Overall, the comprehensive findings suggest that ronment and vegetation in affected areas.
the oil spill had a substantial adverse impact on the
vegetation in the affected area. The lower values of
Author Contributions
NDVI, ARVI, SAVI, GSWIR, and GNIR reflectance
at the oil spill site indicate damage and stress to the J. L.E developed the study, processed the data,
vegetation, while the healthier vegetation and higher and assisted with part of the writing, while E.E.D.
reflectance values at the non-oil spill site serve as a assisted with writing and formatting. The final ver-
reference point for comparison. The correlations be- sion was co-written by all authors.

39
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Conflict of Interest [9] Guyot, G., Baret, F., Jacquemoud, S., 1992. Im-
aging spectroscopy for vegetation studies. Klu-
There is no conflict of interest. wer Academic Publishers: Dordrecht. pp. 145-
165.
References [10] Rowland, E.D., Okpobiri, O., 2021. Floodplain
[1] Okpobiri, O., Harry, A.A., 2022. Monitoring and mapping and risks assessment of the Orashi
detecting the impact of oil sabotage on land us- River using remote sensing and GIS in the Niger
ing multispectral imagery. International Journal Delta Region, Nigeria. Journal of Geographical
of Multidisciplinary Research and Publications Research. 4(2), 10-16.
(IJMRAP). 4(9), 66-74. DOI: https://doi.org/10.30564/jgr.v4i2.3014
[2] Kadafa, A.A., 2012. Oil exploration and spillage [11] Rayment, R.A., 1965. Aspects of the geology
in the Niger Delta of Nigeria. Civil and Envi- of Nigeria—The stratigraphy of the cretaceous
ronmental Research. 2(3), 38-51. and cenozoic deposits. Ibadan University Press:
[3] Okonkwo, C.N.P., Kumar, L., Taylor, S., 2015. Ibadan.
The Niger Delta wetland ecosystem: What [12] Reijers, T., 2011. Stratigraphy and sedimentology
threatens it and why should we protect it? Afri- of the Niger Delta. Geologos. 17(3), 133-162.
can Journal of Environmental Science and Tech- [13] Short, K.C., Stauble, A.J., 1967. Outline of
nology. 9(5), 451-463. geology of Niger Delta. AAPG Bulletin. 51(5),
DOI: https://doi.org/10.5897/AJEST2014.1841 761-779.
[4] Environmental Assessment of Ogoniland [14] Adewoyin, O.O., Joshua, E.O., Akinwumi, I.I.,
[Internet]. Available from: http://www.zaragoza.es/ et al., 2017. Evaluation of geotechnical parame-
contenidos/medioambiente/onu/issue06/1130-eng- ters using geophysical data. Journal of Engineer-
sum.pdf ing and Technological Sciences. 49(1), 95-113.
[5] Oyem, A., (2001). Christian call for action on [15] Etu–Efeotor, J.O., 1997. Fundamentals of petro-
Nigerian oil spill. Sage-Oxford’s Christian En- leum geology. Paragraphic: Port Harcourt.
vironmental Group. [16] Jiang, Z., Huete, A.R., Didan, K., et al., 2008.
[6] Egirani, D.E., Shehata, N., Ugwu, I.M., et al., Development of a two-band enhanced vegeta-
2021. Exposure, geochemical, and spatial dis- tion index without a blue band. Remote Sensing
tribution patterns of an oil spill in parts of the of Environment. 112(10), 3833-3845.
Niger Delta Region of Nigeria. Health and En- DOI: https://doi.org/10.1016/j.rse.2008.06.006
vironment. 2(1), 103-117. [17] Chavez, P.S., 1996. Image-based atmospheric
DOI: https://doi.org/10.25082/HE.2021.01.005 corrections-revisited and improved. Photogram-
[7] Egobueze, F.E., Rowland, E.D., Ebizimo, D.S., metric Engineering and Remote Sensing. 62(9),
2022. Multispectral imagery for detection and 1025-1035.
monitoring of vegetation affected by oil spills [18] Rowland, E.D., Omonefe, F., 2022. Environ-
and migration pattern in Niger Delta Region, mental monitory and impact assessment of solid
Nigeria. World Journal of Advanced Research waste dumpsite using multispectral imagery in
and Reviews. 15(1), 447-458. Yenagoa, Bayelsa state, Nigeria. International
DOI: https://doi.org/10.30574/wjarr.2022.15.1.0682 Journal of Environmental Science and Technol-
[8] Van der Werff, H., Van der Meijde, M., Jansma, ogy. 19(2), 1007-1024.
F., et al., 2008. A spatial-spectral approach for DOI: https://doi.org/10.1007/s13762-021-03456-2
visualization of vegetation stress resulting from [19] Rouse, J.W.Jr., Haas, R.H., Schell, J.A., et al.,
pipeline leakage. Sensors. 8(6), 3733-3743. 1973. Monitoring the Vernal Advancement and
DOI: https://doi.org/10.3390/s8063733 Retrogradation (Green Wave Effect) of Natural

40
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Vegetation [Internet]. Available from: https:// DOI: https://doi.org/10.1016/S0034-4257(96)00072-7

ntrs.nasa.gov/citations/19750020419 [25] Kaufman, Y.J., Tanre, D., 1992. Atmospherically
[20] Running, S.W., Justice, C.O., Salomonson, V., et resistant vegetation index (ARVI) for EOS-MO-
al., 1994. Terrestrial remote sensing science and DIS. IEEE Transactions on Geoscience and Re-
algorithms planned for EOS/MODIS. Interna- mote Sensing. 30(2), 261-270.
tional Journal of Remote Sensing. 15(17), 3587- DOI: https://doi.org/10.1109/36.134076
3620. [26] Sripada, R.P., Heiniger, R.W., White, J.G., et al.,
DOI: https://doi.org/10.1080/01431169408954346 2006. Aerial color infrared photography for de-
[21] Huete, A.R., 1988. A soil-adjusted vegetation termining early in-season nitrogen requirements
index (SAVI). Remote Sensing of Environment. in corn. Agronomy Journal. 98(4), 968-977.
25(3), 295-309. DOI: https://doi.org/10.2134/agronj2005.0200
[22] Huete, A.R., Hua, G., Qi, J., et al., 1992. Nor- [27] Adamu, B., Tansey, K., Ogutu, B., 2015. Using
malization of multidirectional red and NIR vegetation spectral indices to detect oil pollu-
reflectances with the SAVI. Remote Sensing of tion in the Niger Delta. Remote Sensing Letters.
Environment. 41(2-3), 143-154. 6(2), 145-154.
DOI: https://doi.org/10.1016/0034-4257(92)90074-T DOI: https://doi.org/10.1080/2150704x.2015.1015656
[23] Rondeaux, G., Steven, M., Baret, F., 1996. Op- [28] Herrmann, I., Karnieli, A., Bonfil, D.J., et al.,
timization of soil-adjusted vegetation indices. 2010. SWIR-based spectral indices for assessing
Remote Sensing of Environment. 55(2), 95-107. nitrogen content in potato fields. International
DOI: https://doi.org/10.1016/0034-4257(95) Journal of Remote Sensing. 31(19), 5127-5143.
00186-7v DOI: https://doi.org/10.1080/01431160903283892
[24] Gitelson, A.A., Kaufman, Y.J., Merzlyak, M.N., [29] Karnieli, A., Kaufman, Y.J., Remer, L., et al.,
1996. Use of a green channel in remote sensing 2001. AFRI—Aerosol free vegetation index.
of global vegetation from EOS-MODIS. Remote Remote Sensing of Environment. 77(1), 10-21.
sensing of Environment. 58(3), 289-298. DOI: https://doi.org/10.1016/S0034-4257(01)00190-0

41
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Journal of Geographical Research

https://journals.bilpubgroup.com/index.php/jgr

ARTICLE

Considering Regional Connectivity and Policy Factors in the

Simulation of Land Use Change in New Areas: A Case Study of
Nansha New District, China
Zehuan Zheng, Shi Xian*

School of Geographical Sciences and Remote Sensing, Guangzhou University-The University of Birmingham Joint
Research Centre of Service Industries and Urban Development, Guangzhou University, Guangzhou, Guangdong,
510006, China

ABSTRACT
Numerous emerging development areas worldwide are receiving attention; however, current research on land use
change simulation primarily concentrates on cities, urban clusters, or larger scales. Moreover, there is a limited focus
on understanding the impact of regional connectivity with surrounding cities and policy factors on land use change
in these new areas. In this context, the present study utilizes a cellular automata (CA) model to investigate land
use changes in the case of Nansha New District in Guangzhou, China. Three scenarios are examined, emphasizing
conventional locational factors, policy considerations, and the influence of regional connectivity with surrounding
cities. The results reveal several key findings: (1) Between 2015 and 2021, Nansha New District experienced
significant land use changes, with the most notable shifts observed in cultivated land, water area, and construction land.
(2) The comprehensive scenario exhibited the highest simulation accuracy, indicating that Nansha New District, as an
emerging area, is notably influenced by policy factors and regional connectivity with surrounding cities. (3) Predictions
for land use changes in Nansha by 2030, based on the scenario with the highest level of simulation accuracy, suggest
an increase in the proportion of cultivated and forest land areas, alongside a decrease in the proportion of construction
land and water area. This study contributes valuable insights to relevant studies and policymakers alike.
Keywords: CA model; Land use change simulation; Nansha New District

*CORRESPONDING AUTHOR:
Shi Xian, School of Geographical Sciences and Remote Sensing, Guangzhou University-The University of Birmingham Joint Research Centre of
Service Industries and Urban Development, Guangzhou University, Guangzhou, Guangdong, 510006, China; Email: [email protected]
ARTICLE INFO
Received: 30 June 2023 | Revised: 28 July 2023 | Accepted: 4 August 2023 | Published Online: 15 August 2023
DOI: https://doi.org/10.30564/jgr.v6i3.5814
CITATION
Zheng, Z.H., Xian, S., 2023. Considering Regional Connectivity and Policy Factors in the Simulation of Land Use Change in New Areas: A Case
Study of Nansha New District, China. Journal of Geographical Research. 6(3): 42-60. DOI: https://doi.org/10.30564/jgr.v6i3.5814
COPYRIGHT
Copyright © 2023 by the author(s). Published by Bilingual Publishing Group. This is an open access article under the Creative Commons Attribu-
tion-NonCommercial 4.0 International (CC BY-NC 4.0) License. (https://creativecommons.org/licenses/by-nc/4.0/).

42
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

1. Introduction complex system theory progresses and system

simulation platforms are developed, the advantages
Land use/land cover change (LUCC) is a signif- of complex system simulation methods in land use
icant consequence of human-induced alterations to
change research become increasingly apparent [10].
the natural environment, carrying profound impli-
Cellular automata (CA) has emerged as one of the
cations for global environmental change [1]. Conse-
most dominant models for land use simulation due
quently, understanding and predicting LUCC have
to its remarkable ability to capture the interplay
become crucial focal points of research. Particularly
between natural and human-driven factors [1,5,11].
in developing countries, rapid urbanization [2] has led
However, despite the prominence of CA models,
to dramatic land use changes, intensifying the con-
existing studies on land use and land cover change
flict between human activities and land resources.
(LUCC) have primarily focused on cities, urban
China, since its reforms and opening up, has experi-
agglomerations, or larger scales, with relatively
enced rapid urbanization and industrialization, posi-
fewer investigations specifically examining land
tioning the Guangdong-Hong Kong-Macao Greater
use simulation within new areas [3,7,8,12,13]. The rapid
Bay Area as the world’s fourth-largest bay area [3].
and intense land use changes often observed in new
Against this backdrop, the release of the Chinese
areas within cities necessitate close monitoring and
State Council’s “Overall Plan for Guangzhou
thorough understanding. When utilizing CA models
Nansha to Deepen Comprehensive Cooperation with
for geographic simulation, researchers have found
Guangdong, Hong Kong, and Macao for the World”
that the expansion of urban areas is influenced by a
in 2022 has presented new opportunities. China’s
series of driving factors, including topographic con-
dynamic and swift urbanization has given rise to
numerous emerging development areas, and among ditions, transportation factors, and socio-economic
them, Nansha New District, situated in the heart development [14-16].
of the Greater Bay Area, stands out as one of the Traditional drivers of land use change have been
most representative in recent years. The region has extensively studied, encompassing natural factors,
undergone rapid and substantial land use changes, transportation factors, and location factors [17] .
largely influenced by policies and increased regional Among these, particular attention has been given to
connectivity with surrounding cities. Considering natural factors, including terrain conditions repre-
these factors, Nansha New District serves as a sented by the digital elevation model (DEM) [6,7,9,13,18-21],
typical and ideal case study for simulating land use elevation [10,16,22,23], slope [6,7,10,12,13,16,18,19,21-23], and as-
changes in new development areas. pect [6,7,16,18], among others. Notably, slope stands out
Commonly employed land use simulation mod- as one of the most significant topographic factors
els in existing studies include Markov chain models, influencing urban sprawl [7].
system dynamics models, and CLUE-S (Conver- In land use change studies, traffic factors encom-
sion of land use and its effects at small region extent) pass distance to major roads [13,18,21], motorways [10,18],
models [4-7]. Many of these models are founded on the city center [13], regional centers [21,22], railways [21], and
the principles of the cellular automata (CA) mod- highways [21]. Additionally, researchers contend that
el and the CLUE-S model, which have undergone accessibility factors, including public facilities [18]
f u r t h e r e n h a n c e m e n t s a n d r e f i n e m e n t s [8]. and industrial company density [18,21], play a pivot-
One notable advancement is the future land use al role in shaping land use change. Moreover, both
simulation (FLUS) model developed by Liu, which population density and the presence of public facil-
effectively addresses uncertainties and complexities ities [21,24] are acknowledged as influential factors in
associated with interconversions between different land use change dynamics. Furthermore, economic
land use types, demonstrating strong predictive and social development factors are widely employed
capabilities in simulating land use patterns [9]. As in land use change and simulation studies. These

43
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

factors consist of points of interest (POI) data [6,16,18,22], cy factors on land use change in new areas [2,6].
nighttime lighting (NTL) data [18], GDP [6,18-20], economic Based on this premise, the present study fo-
development indicators [18], population statistics [6,19,23,25], cuses on the Guangzhou Nansha New District
and population density metrics [13,16,18,20,23]. within the context of the rapid development of
Scholars are increasingly recognizing the critical the Guangdong-Hong Kong-Macao Greater Bay
role of policy factors in comprehending land use Area. Land cover data from 2015, 2018, and 2021
change and simulation [6,25], especially in regions are utilized to construct three distinct scenarios
significantly influenced by such policies. Sarah in the cellular automata (CA) model: Scenario A
Hasan et al., for instance, highlights the significance considers conventional locational factors, scenario
of diverse planning policy guidelines and regulations A + B incorporates policy factors, and scenario
in shaping urban development [7]. In a separate A + B+ C takes into account the influence of regional
study, Zhipeng Lai et al. utilizes two indicators, connectivity with surrounding cities on the study
namely the Prime Farmland Protection Area and region. The research carefully analyzes the patterns
Ecological Sensitive Area, to assess the impact and characteristics of land use changes in Nansha
of spatially restricted zones on land use change [18]. New District and subsequently employs the scenario
Nonetheless, further explicit research is essential that yields the highest accuracy of simulation to
to thoroughly investigate the potential influence of project its land use data for 2030. By investigating
government policies on land use [25,26]. the patterns and influencing factors of land use
Furthermore, there is a growing recognition of the changes in new areas, this study not only enhances
substantial significance of inter-regional connectivity our understanding but also provides empirical
in urban development and land use change. Hence, support for government policymaking, facilitating
it becomes imperative to investigate the impact of more sustainable land use planning in such regions.
regional connectivity. Particularly, regional connec-
tivity with surrounding cities acts as a quantitative
2. Research data and methods
indicator to assess the degree and quality of inter-
connectedness between a city and other surrounding 2.1 The study area and research data
cities [27,28], facilitating an analysis of the intercon-
nectivity and interdependence among these urban In 2019, with the establishment of the Guang-
centers. Moreover, by considering the total connec- dong-Hong Kong-Macao Greater Bay Area as a na-
tivity between cities [28-30], we can gauge the overall tional strategic zone, Guangzhou Nansha New Dis-
importance of a city network. trict assumed a central and distinctive role as the sole
In summary, it is evident that the development of national new district and comprehensive cooperation
the Greater Bay Area is significantly shaped by both best practice zone in Guangdong Province (Figure
policies and regional connectivity with surrounding 1). This strategic positioning placed Nansha New
cities. Therefore, conducting a comprehensive anal- District at the geographical heart of the development,
ysis of land use changes in new areas, considering resulting in accelerated growth and significant land
the influence of policy factors and inter-regional use changes. As a consequence, the region serves as
connectivity, becomes of paramount importance in an ideal research case for examining land use chang-
understanding the dynamics of land cover changes es within new areas, particularly in the context of
and predicting future developments. However, it is regional connectivity with surrounding cities and the
important to note that there is a notable dearth of influence of policy-driven factors.
research specifically focusing on the influence of re- The data utilized in this study consisted of five
gional connectivity with surrounding cities and poli- main components, as presented in Table 1. First-

44
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 1. Location of Nansha New District.

Table 1. Data utilized in the study and the data source.

Setting of
Classification Data type Year Data source
scenario driving factors
2015,
https://doi.org/10.5281/
Land use data Land use vector data 2018,
zenodo.5816591
2021
Data on natural Resource and Environmental
environmental DEM (including Slope) 2018 Science Data Center Network
factors (http://www.resdc.cn/)
Scenario Metro stations, regional
A centers, urban primary Guangzhou Municipal Bureau of
Scenario Traffic elements
roads, urban secondary 2018 Planning and Natural Resources
A+B and location data
roads, provincial roads and (http://ghzyj.gz.gov.cn/)
highways
The overall land use Guangzhou Planning and Natural
Policy planning 2006
planning for Nansha New Resources Bureau, Nansha Branch
documents -2020
District (2006-2020). (http://www.gzns.gov.cn/)
Economic activity Guangzhou City, Foshan City,
Scenario indicators: total fixed assets, Shenzhen City, Zhongshan City,
A+B+C Socio-economic total production assets; static Dongguan City, Zhuhai City Bureau
2018
indicators population indicators: total of Statistics (https://lwzb.gzstats.
population at the end of the gov.cn:20001/datav/admin/home/
year, etc. www_nj/)
Geographical scale
Guangzhou, Foshan, Shenzhen,
indicators: area, population,
Zhongshan, Dongguan, Zhuhai
etc.; Location indicators:
Data on Bureau of Statistics, Guangzhou
i.e. time and distance to the
indicators Nansha New District Bureau of
center of a large city; Other 2018
related to city Statistics (http://www.gzns.gov.cn/),
indicators: road passenger
connectedness Guangzhou Planning and Natural
data, rail passenger data,
Resources Bureau, Nansha Branch
frequency of place name
(http://www.gzns.gov.cn/)
buzzword searches, etc.

45
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

ly, the land use dataa for Nansha New District in 2.2 Research methods and data processing
Guangzhou City was collected for the years 2015,
Based on the land use data of Nansha New
2018, and 2021. The land use types in Nansha New
District for 2015, 2018, and 2021, this study employs
District were classified into five major categories,
the rate of change of land use types to quantify the
namely cultivated land, forest land, grassland, wa-
extent of changes associated with specific land use
ter area, and construction land, in accordance with
types within the designated study area over the
the national-level classification standard outlined in
specified time frame [32]. The rate of change can
the Current Land Use Classification (GB/T21010-
take on positive or negative values, reflecting the
2017) by the Ministry of Land and Resources [31].
corresponding increase or decrease in the land area
Secondly, the dataset included information on
occupied by the specific land use type, respectively.
traffic elements and location data, encompassing
metro stations, regional centers, urban primary Simulation of land use change trends based on
roads, urban secondary roads, provincial roads, CA model
and highways. The European distance tool in The cellular automaton (CA) model is a powerful
ArcMap software was employed to calculate the bottom-up approach known for its robust spatial
shortest distance from each image element to the computing capability. Recent research on CA
corresponding traffic element within Nansha New applications has demonstrated its ability to simulate
District. Subsequently, the traffic conditions were complex systems through the utilization of simple
normalized using the raster calculator. Thirdly, the local transformation rules [5,33]. In this study, an
data encompassed natural environment factors, extended exploratory investigation was conducted
including a digital elevation model (DEM) with on the definition of CA, particularly focusing on
a resolution of 30 m, and slope data extracted neighborhood configuration and land use conversion
from DEM. Fourthly, policy factors data involved rules. Subsequently, a land use evolution model was
dividing the study area into four functional areas: constructed by integrating artificial neural network
permitted construction area, conditional construction techniques [5,7]. The CA model was employed in
area, restricted construction area, and prohibited simulating scenarios of urban expansion pattern
construction area. Weights were assigned to these changes, with each scenario incorporating different
functional areas using the analytic hierarchy driving factors. Three distinct scenarios were
process. Lastly, urban connectivity indicators constructed: Scenario A, representing conventional
locational factors; Scenario A + B, incorporating
incorporated various metrics, such as geographical
policy factors; and Scenario A + B + C, considering
scale, location, economic activity, static population,
the influence of regional connectivity of surrounding
and other relevant indicators. Due to limited data
cities on land use changes in the study area. Through
availability regarding the regional connectivity
the analysis of the interrelationships between land
between Nansha New District and surrounding
use change characteristics and the different driving
cities, this study employed the regional connectivity
factors during the years 2015, 2018, and 2021, the
between surrounding cities and Guangzhou City as
CA model yielded valuable insights.
an alternative measure for evaluating the regional
connectivity of surrounding cities to Nansha New The model scenarios and driving factors
District. This study aims to investigate the effects of
multiple factors on land use dynamics in the Nansha
a The source is the first Landsat-based annual land cover product
(CLCD) in China released by Professors Jie Yang and Xin Huang of New District. Three scenarios are examined, and
Wuhan University. The data format is raster data, including nine land cellular automata (CA) models are constructed for
types, namely, Cropland, Forest Shrub, Grassland, Water, Snow/Ice,
Barren, Impervious and Wetland, with a spatial resolution of 30 m. each scenario, incorporating factors sequentially as

46
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

follows: conventional locational factors (Scenario A), statistical processing, analysis, and summarization
policy factors (Scenario A + B), and the influence of of expert opinions obtained through anonymous
regional connectivity with surrounding cities on the solicitation. To determine ecological sensitivity
study area (Scenario A + B + C). The specific setup weights for each factor, AHP is utilized, and the
of model scenarios and the associated driving factors judgment matrix is constructed using the expert
are detailed as follows: grade method to calculate the weights of each
(1) Scenario A: Conventional locational factors functional area.
Considering the land use development character- (3) Scenario A + B + C: Conventional locational
istics of Nansha New District and drawing from rel- factors overlaid with policy factors and surround-
evant studies and available data, the primary driving ing cities influence
factors selected for the cellular automata (CA) model Within the context of urban cooperation in the
in the conventional locational factor scenario were Greater Bay Area, Nansha New District, positioned
topographic conditions and traffic factors. These fac- at the core of this region, is witnessing an increasing
tors encompass DEM data (including slopes extract- level of connectivity with surrounding cities. As a
ed from DEM), distances to metro stations, regional result, the land use changes in Nansha New District
centers, urban primary roads, urban secondary roads, may be influenced by the extent of this connectivity.
provincial roads, and highways (please refer to Ta- Thus, it is crucial to include the regional connectivity
ble 1 for detailed information). of surrounding cities as a driving factor in the
(2) Scenario A + B: Conventional locational fac- cellular automata (CA) model. The interconnection
tors overlaid with policy factors and influence among cities serve as indicators
Against the backdrop of the Guangdong-Hong reflecting the impact of urbanization in peripheral
Kong-Macao Greater Bay Area development, areas surrounding cities. Compared to Scenario A +
policy factors significantly influence the land use B, Scenario A + B + C incorporates the driving factor
changes in Nansha New District. The policy factors of regional connectivity of surrounding cities while
scenario (Scenario A + B) in this study considers maintaining consistency with Scenario A + B in
the impact of functional area categories in planning terms of the remaining driving factors. This addition
policies as a crucial driving factor, while keeping aims to enhance the accuracy of land use simulation
the other driving factors consistent with Scenario A. predictions in Nansha New District.
In existing general urban planning documents, the In this study, due to the availability of data re-
construction land control area is primarily classified garding the regional connectivity of the surrounding
into four types: prohibited construction area, cities in Nansha New District, Guangzhou city was
permitted construction area, restricted construction chosen as the focal point for evaluating the city con-
area, and conditional construction area (please refer nectivity of these surrounding cities. Specifically,
to Figure 5). This research establishes the four when computing the regional connectedness of cities
types of functional areas based on the information neighboring Nansha New District, the study incorpo-
provided in the Nansha District General Land Use rated five adjacent cities, namely Foshan, Shenzhen,
Plan (2006-2020). Zhongshan, Dongguan, and Zhuhai, as factors in the
In Scenario A + B, varying degrees of influence simulation calculation using a cellular automata (CA)
on urban construction land are observed among model. The approach described by Lin et al. [30] was
different functional areas. These degrees of influence adopted in this study to calculate the node strength
are quantified by assigning corresponding weights to of each city (network node) within the city cluster
each functional area, employing the expert grading network, thereby assessing the strength of city link-
method in conjunction with the analytic hierarchy ages using multiple indicators, as outlined in Table
process (AHP). The expert grade method involves 1. In accordance with weighted network theory, the

47
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

city nodal strength is a representation of the sum of weight assigned to the kth method. The sum of all
weights associated with a nodal bond [34-36]: weights, represented by ΣW, is equal to 1.

∑ Accuracy assessment
n
=Si j =1
Cij r (i ≠ j )  (1)
ij

Si represents the node intensity of city i. n rep- To assess the accuracy of model simulations,
resents the total number of cities, Cij represents the previous studies have commonly employed overall
connectivity score between city i and city j, and rij accuracy (OA) and Kappa coefficients [33,40]. Addi-
represents the relationship between city i and city j. tionally, scholars [41] proposed the factor of merit
A value of 1 is assigned to rij if there is a relationship (FoM) as an alternative measure to evaluate sim-
between the two cities, and a value of 0 if there is no ulation accuracy. FoM relies solely on the number
relationship. of varying cells in the simulated process and offers
In this research, the selection of indicators was a more robust assessment of simulation accuracy
[13,16]
guided by domain knowledge and expert experience, . Building upon this, in this study, historical
encompassing several widely utilized socioeconomic data was used for simulation, with the 2018 dataset
indicators, namely gross economic product, fixed as- serving as the initial input and the 2021 dataset used
set investment, and year-end resident population (as for verifying the validity of the CA model [1]. Kappa
presented in Table 1). A higher average coefficient coefficient, overall accuracy (OA), and FoM were
of determination generally indicates a more reliable employed to evaluate the accuracy of the model. The
dataset [37,38]. To effectively integrate the diverse data Kappa coefficient and OA values range from 0 to 1,
sources, an intelligent optimization algorithm known with higher values indicating greater classification
as the selective weighted combination method was accuracy achieved by the model. A Kappa coefficient
employed in this study. Specifically, we utilized the above 0.80 is typically considered indicative of very
well-established genetic algorithm proposed by Zhou high simulation accuracy [4,6,9]. As for FoM, values
et al. [39] for optimization purposes. This algorithm exceeding 0.2 are suggested to indicate an extremely
enabled us to obtain optimal weights and determine usable model [21,22,42].
the coefficients of determination between different
socioeconomic indicators. The adaptability function 3. Simulation results and analysis
of the genetic algorithm was calculated as follows:
3.1 Characteristics and trend analysis of land
f ( x) = ∑i =1 ri / n (2)
n

use change
ri represents the coefficient of determination be-
tween the weighted result and the ith socioeconomic As discussed in the methodology section, the
indicator. n is the total number of indicators. changes in land area of different land types in Nan-
The final outcome reveals a significantly high- sha New District were first calculated for the period
er urban connectivity intensity when compared to from 2015 to 2021 (see Figure 2).
any individual data source. The urban connectivity The findings of the study revealed several note-
values between Nansha New District and the sur- worthy trends (see Figure 2). Firstly, there was a
rounding cities are presented in Figure 6. The final gradual decrease in the area of cultivated land, forest
connectivity score (Cij) between city i and city j can land, and water area in Nansha New District over
be expressed as follows: the specified period. Cultivated land accounted for
more than half of the total area during 2015-2021
Cij = ∑k =1Wk Cijk (3)
t

and served as the predominant land type in the dis-

t is the total number of methods. Cijk represents trict. However, its proportion experienced a year-
the result obtained from the kth method for the jth on-year decline, with the proportion in 2021 being
observation of the ith variable. Wk represents the approximately 1.5 percentage points lower than that

48
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

study employs a land use transition matrix to depict

the interconversion of land use types in the research
area (see Figure 3 and Table 2). The findings indi-
cate that the conversion of land to other use types
during the 2015-2021 period is relatively limited.
Specifically, between 2015 and 2018, the most sig-
nificant land use conversion occurred from cultivated
land to other use types, followed by water area. Al-
most half of the grassland area was transformed into
construction land, and forest land was predominantly
converted to cultivated land, with over 10 percent of
the total forest area transitioning to cultivated land,
and only a minimal proportion transitioning to con-
struction land. Water area conversions were primar-
ily towards cultivated land, while construction land
conversions were limited, not exceeding 0.2 percent.
Figure 2. Land use changes in Nansha New District, Guangzhou During the 2018-2021 period, the land use conver-
(2015-2021). sions remained consistent with those observed in the
in 2015. The proportion of forest land also exhibited 2015-2018 period. Notably, there was a significant
a decreasing trend during the same period. Addition- conversion of construction land to cultivated land.
ally, the water area saw a reduction between 2015 Furthermore, the proportion of water area transition-
and 2021, accounting for more than one-fifth of the ing to cultivated land decreased from nearly 7.43
total area in 2015 and less than one-fifth in 2021. In percent during 2015-2018 to nearly 4.52 percent
contrast, the area of construction land exhibited a during 2018-2021, whereas the proportion of water
steady increase, while the area of grassland initially area transitioning to construction land increased
decreased and then showed an upward trend. Con- from about 2.23 percent during 2015-2018 to about
struction land accounted for less than one-fifth of the 3.45 percent during 2018-2021. The examination of
total area in 2015 and surpassed one-fourth in 2021, the land use transition matrix for 2015-2021 further
representing more than one-fourth of the overall supports the aforementioned results, indicating that
growth between 2015 and 2021. In 2015, construc- the conversion of larger land areas to construction
tion land was primarily concentrated in the eastern land predominantly affected cultivated land and wa-
and western townships, but it gradually expanded ter area. Specifically, during the period 2015-2021,
towards the central and northern areas near the about 7.64 percent of the total area of cultivated
center of Guangzhou city after 2015. The proportion land was converted to construction land. Similarly,
of grassland area demonstrated a decrease followed roughly 5.56 percent of the water area underwent a
by an increase during the period from 2015 to 2021. transition to construction land. It is worth noting that
During the development and construction of Nan- this study also reveals a trend of gradual increase in
sha New District from 2015 to 2021, a considerable the proportion of cultivated land and water area tran-
amount of cultivated land and water area were con- sitioning to construction land. These findings sub-
verted to meet the demands of urban construction. stantiate the negative impact of urban construction
To illustrate the changes in land use types within demand on cultivated land and water area during the
the Nansha New District from 2015 to 2021, this 2015-2021 period.

49
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 3. Transition matrix of land use type in Nansha New District from 2015 to 2021.

Table 2. Transition matrix of land use type in Nansha New District from 2015 to 2021.

2018
Land use type Grassland Cultivated Construction Forest land Water area Total
(km2) land (km2) land (km2) (km2) (km2) (km2)
Grassland (km2) 0.06 0.00 0.06 0.00 0.12
2
Cultivated land (km ) 0.01 301.33 11.11 0.22 1.37 314.03
Construction land (km2) 115.37 0.22 115.58
2
2015 Forest land (km ) 2.54 0.07 20.97 23.58
2
Water area (km ) 0.01 9.90 2.97 0.00 120.29 133.18
2
Total (km ) 0.07 313.78 129.58 21.19 121.88 586.50
Land use dynamic (%) –14.22 –0.03 4.04 –3.38 –2.83
2021
Grassland (km2) 0.05 0.00 0.02 0.00 0.07
2
Cultivated land (km ) 0.01 298.83 13.18 0.74 1.01 313.78
2
Construction land (km ) 0.02 129.40 0.16 129.58
2018 2
Forest land (km ) 0.90 0.05 20.24 21.19
Water area (km2) 0.04 5.51 4.21 0.01 112.11 121.88
2
Total (km ) 0.10 305.25 146.86 20.99 113.28 586.50
Land use dynamic (%) 15.81 -0.91 4.44 –0.31 –2.35
2021
2
Grassland (km ) 0.04 0.01 0.08 0.00 0.12
2
Cultivated land (km ) 0.02 287.24 23.99 0.61 2.17 314.03
Construction land (km2) 0.02 115.25 0.32 115.58
2
2015 Forest land (km ) 3.06 0.14 20.37 0.01 23.58
2
Water area (km ) 0.05 14.93 7.41 0.01 110.78 133.18
2
Total (km ) 0.10 305.25 146.86 20.99 113.28 586.50
Land use dynamic (%) –2.57 –0.47 4.51 –1.83 –2.49

50
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

3.2 The calculated results for the main driv- 0.242, 0.145, and 0.075, respectively (refer to Figure
ing factors in the three scenarios 5). Among the functional areas considered, the per-
mitted construction area exhibited the highest weight
In this study, three distinct scenarios were consid- value, signifying its significant influence on the
ered in developing the CA models, and the calcula- allocation of construction land within Nansha New
tion of the main driving factors for each scenario is District. Conversely, the prohibited construction area
presented as follows: had the lowest weight value, indicating its minimal
The calculated results for the main driving fac- impact on the dynamics of construction land within
tors in Scenario A Nansha New District.

In Scenario A, the driving factors for the CA The calculated results for the main driving fac-
model consisted of conventional locational factors, tors in Scenario A + B + C
including topographic conditions and traffic factors, Based on the data provided in Figure 6, the re-
with the normalized values presented in Figure 4. sults indicate that Nansha New District displayed
the highest level of connectivity with Shenzhen, sur-
The calculated results for the main driving fac-
passing a threshold of 0.6. Furthermore, Foshan City
tors in Scenario A + B
exhibited a connectivity degree exceeding 0.5. In
The calculation results, derived from the imple- contrast, Zhongshan, Dongguan, and Zhuhai demon-
mentation of the expert grade method and analytic strated comparatively lower levels of connectivity
hierarchy process (AHP), demonstrated that the with Nansha New District.
weights assigned to the permitted construction area, By utilizing the Euclidean distance tool, this
conditional construction area, restricted construction study calculated the distances between the cell
area, and prohibited construction area were 0.538, centers of Nansha New District and its surrounding

Figure 4. Driving factors used for scenario A: (I) DEM; (II) slope; (III) distance to metro stations; (IV) distance to regional centers; (V)
distance to urban primary roads; (VI) distance to urban secondary roads; (VII) distance to provincial roads; (VIII) distance to highways.

51
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 5. Driving factors used for scenario A + B: (Ⅰ) Planning map of different functional areas in Nansha New District; (Ⅱ) Degree
of land use impact of different functional areas.

Figure 6. Driving factors used for scenario A + B + C: (I-Ⅵ indicate Dongguan City, Zhuhai City, Zhongshan City, Shenzhen City,
Foshan City and location of surrounding cities, respectively).

52
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

cities. These distances were then combined with the

level of city connection to generate Figure 6.
The results provide the following insights:
Industrial Development Structure: The land use
patterns in Nansha New District exhibit a stronger
influence from areas that closely align with its indus-
trial structure.
Economic Development Level: The land use in
Nansha New District is more influenced by cities
with a stronger economic base. These cities act as
development poles, facilitating the exchange of vari-
ous factors such as human flow, logistics, and capital
flow with neighboring urban clusters, fostering posi-
tive interactions.
Transportation Accessibility and Population Mo-
bility: Areas with higher levels of traffic accessibil-
ity and greater population mobility exert a stronger
influence on the land use patterns of Nansha New
District.

3.3 Simulation results of land use changes

using the CA Model and the accuracy assess-
ment

The land use change simulation in Nansha New

District in 2021 was conducted using the cellular
automata (CA) model, with the 2018 land use data
serving as the baseline. Three distinct scenarios,
namely A, A + B, and A + B + C, were established
to simulate the land use patterns. These scenarios
incorporated conventional locational factors (Scenar-
io A), policy factors (Scenario A + B), and regional
connectivity with surrounding cities (Scenario A +
B + C). The simulation results depicting the land use
changes in Nansha New District in 2021 under the
three different scenarios are presented in Figure 7
and Table 3.
In this study, the processed data were utilized as
input for the CA model to simulate and predict the
land use change trend of Nansha New District from
2015 to 2021. The findings (refer to Figure 7 and
Table 3) revealed that the highest overall accuracy
(OA) of the model was achieved when considering
the overlay of conventional locational factors, relat- Figure 7. Simulated land use in Nansha New District, Guang-
ed policy factors, and influence factors of regional zhou in 2021 under scenario A; A + B; A + B + C.

53
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 3. Simulation results of land use changes and the accuracy assessment.
Scenario Type A A+B A+B+C
Area Percentage Area Percentage Area Percentage
(km2) (%) (km2) (%) (km2) (%)
Cultivated land 305.25 52.05 305.25 52.05 305.25 52.05
Forest land 20.99 3.58 21.02 3.58 21.03 3.59
Grassland 0.08 0.01 0.08 0.01 0.07 0.01
Water area 113.31 19.32 113.28 19.31 113.28 19.31
Construction land 146.86 25.04 146.86 25.04 146.86 25.04
Accuracy of land use types in Nansha New District in 2021 for three scenarios
Scenario type A A+B A+B+C
Overall accuracy (OA) 0.8933 0.898 0.899
Kappa coefficient 0.8304 0.8372 0.8394
FoM accuracy value 0.0554 0.0565 0.0516

connectivity with surrounding cities simultaneously. factors, policy factors, and the influence of regional
This scenario yielded an OA of 0.8990, a Kappa connectivity with surrounding cities, the proportion
coefficient of 0.8394, and a factor of merit (FoM) of of forest land area was the highest, reaching 3.5858%
0.0516. The changes in OA values and Kappa coeffi- of the total land area. In contrast, the area occupied
cients were more pronounced in Scenario A+B com- by grassland was the smallest, representing only
pared to Scenario A, suggesting that policy factors 0.0114% of the land. Regarding water area, the larg-
exerted a significant influence on land use change in est proportion was observed when solely considering
Nansha New District. However, the changes in OA the conventional location factors, accounting for
values and Kappa coefficients in Scenario A + B + 19.3197% of the total land area.
C showed subtle variations, which were not as sig-
nificant as the changes from Scenario A to Scenario 3.4 Future land use prediction of Nansha New
A + B, indicating that the influence of regional con- District in 2030
nectivity of surrounding cities on land use change in
Nansha New District was not as prominent as that Based on the significant findings mentioned
of policy factors. Nevertheless, the combined OA above, Scenario A + B + C, which encompassed the
values and Kappa coefficients indicated that the sim- integration of conventional locational factors, policy
ulation accuracy of Scenario A + B + C considering factors, and the influence of regional connectivity
multiple driving factors was the highest (refer to Ta- with surrounding cities, exhibited the highest accu-
ble 3). racy in the model simulation. Consequently, this sce-
In this study, we examined the impact of policy nario was chosen for predicting the characteristics of
factors and regional connectivity with surrounding land use change in Nansha New District for the year
cities on the land use dynamics of Nansha New Dis- 2030. Using the CA model, the study simulated and
trict, a burgeoning development area. Our findings predicted the expected land use types in Nansha New
revealed that policy factors played a more prominent District for 2030.
role in shaping land use changes. When analyzing Based on the predicted results (Figure 8), several
the simulated area and proportion of different land notable observations can be made regarding the an-
use types in Nansha New District in 2021 under var- ticipated land use changes in Nansha New District
ious scenarios, significant changes were observed in between 2021 and 2030. Firstly, the proportion of
forest land, grassland, and water area. Notably, when cultivated land area is projected to increase from
considering the overlay of conventional locational 52.05% in 2021 to 53.24% in 2030. Secondly, there

54
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 8. Land use predicted results of Nansha New District in 2030.

is a minor decrease expected in the proportion of wa- lating land use change within newly emerging areas.
ter area, declining from 19.31% in 2021 to 18.95% Moreover, there is a notable scarcity of studies that
in 2030. Thirdly, the proportion of construction simultaneously consider conventional locational
land area is predicted to decline, decreasing from factors, the influence of regional connectivity with
25.04% in 2021 to 24.2% in 2030. Additionally, the surrounding cities, and policy factors on land use
proportion of forest land area is projected to exhibit change within these new areas. This paper aims to
a marginal increase, rising from 3.58% in 2021 to address and critically discuss the limitations identi-
3.59% in 2030. Lastly, the proportion of grassland fied in the aforementioned relevant studies. Utilizing
area is anticipated to remain relatively stable, with a case study of Nansha New District in China and
no significant changes expected during the specified employing the CA model, this study simulated the
period.
land use changes in Nansha New District for the
years 2015, 2018, and 2021, while also predicting its
4. Conclusions and discussion changes in 2030. The CA model was constructed un-
The rapid and profound urbanization in China der three scenarios: conventional locational factors
has given rise to numerous new development areas. alone, conventional locational factors combined with
However, existing research primarily focuses on policy factors, and conventional locational factors
land use change simulation at the city, urban cluster, integrated with policy factors and regional connec-
or larger scales, offering limited insights into simu- tivity of the surrounding cities. The latter two scenar-

55
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

ios were specifically designed to elucidate the impact coordination between human activities and land uti-
of policy factors and regional connectivity on land lization. To address the problem of cultivated land
use changes within Nansha New District. The find- fragmentation, a comprehensive approach focusing
ings revealed that: (1) significant changes in various on the binding role of policies and regulations is
land uses in Nansha New District during the period essential. Timely planning evaluation is crucial, and
from 2015 to 2021, particularly in cultivated land, greater attention should be given to the protection of
water area, and construction land. (2) Comparing the basic farmland and cultivated land. Strictly penaliz-
predicted land use data for 2030, generated from the ing violations of the law involving construction on
comprehensive model with the highest simulation cultivated land should be prioritized. Additionally,
accuracy, with the current land use status in 2021, it leveraging the advantages of regional connectiv-
was observed that the proportion of cultivated land ity with neighboring cities is vital. Mitigating the
and forest land area increased, while the proportion haphazard expansion of urban construction land re-
of construction land and water area decreased. The quires emphasizing compact land use and effective
proportion of grassland area showed insignificant land use planning. Key strategies include promoting
changes. mixed land use, optimizing the integration of above
By examining the land use data from 2015, 2018, and below-ground space, and exploring untapped
and 2021, along with the predicted land use data for urban land potential. Innovative modes of land use
2030, several key observations emerge regarding the planning and management, including evaluating the
land use dynamics in Nansha New District. Firstly, alignment between planning policies and actual site
the dynamic development of new areas often leads to layout, can enhance the effectiveness of planning de-
rapid and significant land use transformations with- cisions.
in relatively small areas. Secondly, cultivated land This study provides support for the argument
holds a prominent proportion in the current land use that policy factors and regional connectivity with
pattern; however, it is currently facing challenges surrounding cities exert a substantial influence on
such as gradual fragmentation and declining intensi- the development of new areas. Three scenarios were
fication. This fragmentation and declining productiv- investigated in this research, each considering dif-
ity pose significant obstacles to agricultural activities ferent influential factors. The findings reveal that the
in Nansha New District, highlighting the impact of scenario achieving the highest overall model accu-
urbanization on agricultural land. Furthermore, the racy was the comprehensive model, which simulta-
expansion of construction land is noteworthy, par- neously considers conventional locational factors,
ticularly at the expense of cultivated land. The cur- relevant policy factors, and the influence of regional
rent distribution of construction land exhibits frag- connectivity with surrounding cities. This indicates
mentation and disorderly spreading, particularly in that policy factors and regional connectivity with
the eastern part of the district. The predicted spatial surrounding cities significantly impact the land use
distribution of construction land in 2030, especially change in Nansha New District as an emerging area.
in relation to the four industrial zones highlighted in The study holds noteworthy theoretical and empir-
the Urban Development Plan of Nansha New Dis- ical significance, contributing to a comprehensive
trict, requires more attention in future planning eval- understanding of the driving forces behind land use
uations. changes within flourishing new areas, especially in
In the future development of Nansha New Dis- the context of regional association and cooperation
trict, to formulate policies that effectively address strategies.
land-use challenges arising from rapid development, Admittedly, this study selected only a limited
decision-makers should promote high-quality devel- number of factors as conventional locational fac-
opment in Nansha New District through enhanced tors, and the approximation of the connectivity of

56
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

surrounding cities to Nansha New District based on lar automata. International Journal of Geograph-
their connectivity to Guangzhou might introduce ical Information Science. 24(5), 783-802.
some inaccuracies in the results. Additionally, the DOI: https://doi.org/10.1080/13658810903270551
consideration of only functional areas in the CA [3] Zhou, R., Lin, M., Gong, J., et al., 2019. Spatio-
model when analyzing policy factors might be fur- temporal heterogeneity and influencing mech-
ther improved by incorporating additional factors anism of ecosystem services in the Pearl River
for a more comprehensive analysis to enhance sim- Delta from the perspective of LUCC. Journal of
ulation accuracy in future studies. Nevertheless, this Geographical Sciences. 29(5), 831-845.
research may offer valuable implications for simu- DOI: https://doi.org/10.1007/s11442-019-1631-0
lating land use change in similar rapidly developing [4] Beroho, M., Briak, H., Cherif, E.K., et al., 2023.
new areas, and it provides valuable insights for poli- Future scenarios of land use/land cover (LULC)
cymakers. based on a CA-markov simulation model: Case
of a Mediterranean watershed in Morocco. Re-
Author Contributions mote Sensing. 15(4), 1162.
DOI: https://doi.org/10.3390/rs15041162
Zehuan Zheng: Conceptualization, methodology, [5] Li, X., Yang, Q., Liu, X., 2007. Genetic algo-
Formal analysis, Data Curation, Writing-Original rithms for determining the parameters of cellular
draft preparation, Writing-Review; Editing. automata in urban simulation. Science in China
Shi Xian: Conceptualization, Writing-Review; Series D: Earth Sciences. 50(12), 1857-1866.
Editing. DOI: https://doi.org/10.1007/s11430-007-0127-4
[6] Lin, W., Sun, Y., Nijhuis, S., et al., 2020. Sce-
Conflict of Interest nario-based flood risk assessment for urbanizing
deltas using future land-use simulation (FLUS):
Declaration of no conflict of interest.
Guangzhou Metropolitan Area as a case study.
Science of the Total Environment. 739, 139899.
Funding DOI: https://doi.org/10.1016/j.scitotenv.2020.139899
This research received was funded by the Na- [7] Hasan, S., Shi, W., Zhu, X., et al., 2020. Fu-
tional Natural Science Foundation of China (Grant ture simulation of land use changes in rapidly
No. 42271217) and the Guangdong Planning Office urbanizing South China based on land change
of Philosophy and Social Science (Grant No.GD- modeler and remote sensing data. Sustainability.
21CGL15). 12(11), 4350.
DOI: https://doi.org/10.3390/su12114350
[8] Li, D., Li, X., Liu, X., et al., 2012. GPU-CA
References
model for large-scale land-use change simula-
[1] Liu, X., Hu, G., Ai, B., et al., 2018. Simulating tion. Chinese Science Bulletin. 57(19), 2442-
urban dynamics in China using a gradient cel- 2452.
lular automata model based on S-shaped curve DOI: https://doi.org/10.1007/s11434-012-5085-3
evolution characteristics. International Journal [9] Liu, X., Liang, X., Li, X., et al., 2017. A future
of Geographical Information Science. 32(1), 73- land use simulation model (FLUS) for simulat-
101. ing multiple land use scenarios by coupling hu-
DOI: https://doi.org/10.1080/13658816.2017.1376065 man and natural effects. Landscape and Urban
[2] Liu, X., Li, X., Shi, X., et al., 2010. Simulating Planning. 168, 94-116.
land-use dynamics under planning policies by DOI: https://doi.org/10.1016/j.landurbplan.2017.09.019
integrating artificial immune systems with cellu- [10] Liu, X., Li, X., Shi, X., et al., 2012. A multi-

57
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

type ant colony optimization (MACO) method [17] Chen, C., Liu, Y., 2021. Spatiotemporal changes
for optimal land use allocation in large areas. In- of ecosystem services value by incorporating
ternational Journal of Geographical Information planning policies: A case of the Pearl River Del-
Science. 26(7), 1325-1343. ta, China. Ecological Modelling. 461, 109777.
DOI: https://doi.org/10.1080/13658816.2011.63 DOI: https://doi.org/10.1016/j.ecolmodel.2021.
5594 109777
[11] van Vliet, J., Hurkens, J., White, R., et al., 2012. [18] Lai, Z., Chen, C., Chen, J., et al., 2022.
An activity-based cellular automaton model to Multi-scenario simulation of land-use change
simulate land-use dynamics. Environment and and delineation of urban growth boundaries in
Planning B: Planning and Design. 39(2), 198- county area: A case study of Xinxing County,
212. Guangdong Province. Land. 11(9), 1598.
DOI: https://doi.org/10.1068/b36015 DOI: https://doi.org/10.3390/land11091598
[12] Xiong, N., Yu, R., Yan, F., et al., 2022. Land use [19] Liao, W., Liu, X., Xu, X., et al., 2020. Projec-
and land cover changes and prediction based tions of land use changes under the plant func-
on multi-scenario simulation: A case study of tional type classification in different SSP-RCP
Qishan County, China. Remote Sensing. 14(16), scenarios in China. Science Bulletin. 65(22),
4041. 1935-1947.
DOI: https://doi.org/10.3390/rs14164041 DOI: https://doi.org/10.1016/j.scib.2020.07.014
[13] Xu, X., Zhang, D., Liu, X., et al., 2022. Simu- [20] Li, L., Huang, X., Yang, H., 2023. Scenar-
lating multiple urban land use changes by inte- io-based urban growth simulation by incorporat-
grating transportation accessibility and a vec- ing ecological-agricultural-urban suitability into
tor-based cellular automata: A case study on city a future land use simulation model. Cities. 137,
of Toronto. Geo-spatial Information Science. 104334.
25(3), 439-456. DOI: https://doi.org/10.1016/j.cities.2023.104334
DOI: https://doi.org/10.1080/10095020.2022.20 [21] Zhai, Y., Yao, Y., Guan, Q., et al., 2020. Sim-
43730 ulating urban land use change by integrating a
[14] Liu, X., Li, X., Shi, X., et al., 2010. Simulating convolutional neural network with vector-based
land-use dynamics under planning policies by cellular automata. International Journal of Geo-
integrating artificial immune systems with cellu- graphical Information Science. 34(7), 1475-
lar automata. International Journal of Geograph- 1499.
ical Information Science. 24(5), 783-802. DOI: https://doi.org/10.1080/13658816.2020.17
DOI: https://doi.org/10.1080/13658810903270551 11915
[15] Wu, F., 2002. Calibration of stochastic cellular [22] Yao, Y., Liu, X., Li, X., et al., 2017. Simulating
automata: The application to rural-urban land urban land-use changes at a large scale by in-
conversions. International Journal of Geographi- tegrating dynamic land parcel subdivision and
cal Information Science. 16(8), 795-818. vector-based cellular automata. International
DOI: https://doi.org/10.1080/13658810210157769 Journal of Geographical Information Science.
[16] Zhang, J., Hou, Y., Dong, Y., et al., 2022. Land 31(12), 2452-2479.
use change simulation in rapid urbanizing re- DOI: https://doi.org/10.1080/13658816.2017.13
gions: a case study of Wuhan urban areas. Inter- 60494
national Journal of Environmental Research and [23] Liu, L., Yu, S., Zhang, H., et al., 2023. Analysis
Public Health. 19(14), 8785. of land use change drivers and simulation of dif-
DOI: https://doi.org/10.3390/ijerph19148785 ferent future scenarios: Taking Shanxi Province

58
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

of China as an example. International Journal 63302

of Environmental Research and Public Health. [31] GB/T 21010-2017 (GBT21010-2017). Current
20(2), 1626. land use classification [Internet]. China National
DOI: https://doi.org/10.3390/ijerph20021626 Standardization Administration. Available from:
[24] Zhang, D., Liu, X., Wu, X., et al., 2019. Multi- https://www.chinesestandard.net/PDF/English.
ple intra-urban land use simulations and driving aspx/GBT21010-2017
factors analysis: a case study in Huicheng, Chi- [32] Yan, J., Zhang, Y., Liu, L., et al., 2002. Land use
na. GIScience & Remote Sensing. 56(2), 282- and landscape pattern change: A linkage to the
308. construction of the Qinghai-Xizang Highway.
DOI: https://doi.org/10.1080/15481603.2018.15 Journal of Geographical Sciences. 12(3), 253-
07074 265.
[25] Arficho, M., Thiel, A., 2020. Does land-use pol- DOI: https://doi.org/10.1007/bf02837543
icy moderate impacts of climate anomalies on [33] Liu, X., Li, X., Anthony, G.O.Y., 2006. Multi-
lulc change in dry-lands? An empirical enquiry agent systems for simulating spatial decision
into drivers and moderators of LULC change in behaviors and land-use dynamics. Science in
Southern Ethiopia. Sustainability. 12(15), 6261.
China Series D: Earth Sciences. 49(11), 1184-
DOI: https://doi.org/10.3390/su12156261
1194.
[26] Kolb, M., Galicia, L., 2018. Scenarios and story
DOI: https://doi.org/10.1007/s11430-006-1184-9
lines: Drivers of land use change in southern
[34] Barrat, A., Barthelemy, M., Pastor-Satorras, R., et
Mexico. Environment, Development and Sus-
al., 2004. The architecture of complex weighted
tainability. 20(2), 681-702.
networks. Proceedings of the National Academy of
DOI: https://doi.org/10.1007/s10668-016-9905-5
Sciences of the United States of America. 101(11),
[27] Fan, T., 2006. Improvements in intra-European
3747-3752.
inter-city flight connectivity: 1996-2004. Journal
DOI: https://doi.org/10.1073/pnas.0400087101
of Transport Geography. 14(4), 273-286.
[35] Batty, M., 2018. The new science of cities. The
DOI: https://doi.org/10.1016/j.jtrangeo.2005.
MIT Press: Cambridge.
08.006
[36] Newman, M.E., 2004. Analysis of weighted net-
[28] Allard, R.F., Moura, F., 2016. The incorpora-
tion of passenger connectivity and intermodal works. Physical Review E. 70(5), 056131.
considerations in intercity transport planning. DOI: https://doi.org/10.1103/PhysRevE.70.056131
Transport Reviews. 36(2), 251-277. [37] Onnela, J.P., Saramäki, J., Hyvönen, J., et al.,
DOI: https://doi.org/10.1080/01441647.2015.10 2007. Analysis of a large-scale weighted net-
59379 work of one-to-one human communication. New
[29] Yang, Y., Li, D., Li, X., 2019. Public transport Journal of Physics. 9(6), 179.
connectivity and intercity tourist flows. Journal DOI: https://doi.org/10.1088/1367-2630/9/6/179
of Travel Research. 58(1), 25-41. [38] Opsahl, T., Agneessens, F., Skvoretz, J., 2010.
DOI: https://doi.org/10.1177/0047287517741997 Node centrality in weighted networks: Gener-
[30] Lin, J., Wu, Z., Li, X., 2019. Measuring in- alizing degree and shortest paths. Social Net-
ter-city connectivity in an urban agglomeration works. 32(3), 245-251.
based on multi-source data. International Jour- DOI: https://doi.org/10.1016/j.socnet.2010.03.006
nal of Geographical Information Science. 33(5), [39] Zhou, Z.H., Wu, J., Tang, W., 2002. Ensembling
1062-1081. neural networks: Many could be better than all.
DOI: https://doi.org/10.1080/13658816.2018.15 Artificial Intelligence. 137(1-2), 239-263.

59
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

DOI: https://doi.org/10.1016/S0004-3702(02) et al., 2007. Accuracy assessment for a simula-

00190-X tion model of Amazonian deforestation. Annals
[40] Li, X., Yeh, A.G.O., 2002. Neural-network-based of the Association of American Geographers.
cellular automata for simulating multiple land 97(4), 677-695.
use changes using GIS. International Journal of DOI: https://doi.org/10.1111/j.1467-8306.2007.
Geographical Information Science. 16(4), 323- 00577.x
343. [42] Gerard, B.M.H., 1998. Error propagation in en-
DOI: https://doi.org/10.1080/13658810210137004 vironmental modelling with GIS. CRC Press:
[41] Pontius Jr, R.G., Walker, R., Yao-Kumah, R., Boca Raton.

60
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Journal of Geographical Research

https://journals.bilpubgroup.com/index.php/jgr

ARTICLE

Spatial Agglomeration and Diffusion of Population Based on a Regional

Density Function Approach: A Case Study of Shandong Province in China
Xiaohan Zhao , Yanbin Chen*

College of Geography and Environment, Shandong Normal University, Jinan, Shandong, 250358, China

ABSTRACT
Population density functions have long been used to describe the spatial structure of regional population
distributions. Several studies have been conducted to examine the population distribution in Shandong Province,
China, but few have applied regional density functions to the analysis. Therefore, based on the 2000, 2010, and
2020 population censuses, this study used monocentric and polycentric regional density functions to study the
characteristics of population agglomeration and diffusion in Shandong. This is followed by an in-depth discussion
based on population growth rate data and hot- and cold-spot analyses. The results showed that the Shandong Province
population was spatially unevenly distributed. Population growth rates were higher in urban centers and counties,
with more significant changes in population size in the eastern coastal areas than in the inland areas. As verified in this
study, the logarithmic form of the single-center regional density function R2 was greater than 0.8, which was in line
with the population spatial structure of Shandong Province. During the study period, the estimated population density
of the regional center and the absolute value of the regional population density gradient both increased, indicating
a clear and increasing trend of centripetal agglomeration of regional centers over the study period. Overall, the R2
value of the multicenter region density function was higher than that of the single-center region density function. The
polycentric regional density function showed that the population density gradient of some centers had a downward
trend, which reflected the spatial development trend of outward diffusion in these centers. Meanwhile, the variation
in the estimated population density and the population density gradient exhibited differences in the central population
distribution patterns at different levels.
Keywords: Regional density functions; Population spatial structure; Shandong Province

*CORRESPONDING AUTHOR:
Yanbin Chen, College of Geography and Environment, Shandong Normal University, Jinan, Shandong, 250358, China; Email: [email protected]
ARTICLE INFO
Received: 5 July 2023 | Revised: 10 August 2023 | Accepted: 14 August 2023 | Published Online: 18 August 2023
DOI: https://doi.org/10.30564/jgr.v6i3.5826
CITATION
Zhao, X.H., Chen, Y.B., 2023. Spatial Agglomeration and Diffusion of Population Based on a Regional Density Function Approach: A Case Study
of Shandong Province in China. Journal of Geographical Research. 6(3): 61-80. DOI: https://doi.org/10.30564/jgr.v6i3.5826
COPYRIGHT
Copyright © 2023 by the author(s). Published by Bilingual Publishing Group. This is an open access article under the Creative Commons Attribu-
tion-NonCommercial 4.0 International (CC BY-NC 4.0) License. (https://creativecommons.org/licenses/by-nc/4.0/).

61
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

1. Introduction explore the fitting of the regional population density

distribution and analyze the regional spatial struc-
Population is the most dynamic and active ture and its evolutionary characteristics. However,
element in the process of urban development, and
compared to urban population density models, the
its spatial distribution in cities has always been an
research results are still relatively limited. With the
important part of urban geography research [1-3].
innovation of information technology, restructuring
The spatial structure of a population refers to the
of economic space, and development of transport
tendency of the population to agglomerate or diffuse
infrastructure, the trend toward urban polycentricity
within a given region, which can, to a certain extent,
is becoming increasingly evident. The existence of
reveal the strength or weakness of the region’s so-
multiple centers with specialized economic, social,
cioeconomic activities [4]. With the increasingly
and cultural functions, as well as the interaction
significant spatial reconfiguration of the inner
between the centers, makes monocentric models no
city, studying population distribution within the
longer suitable for describing contemporary urban
city has become increasingly vital [5]. Population
structures [12]. Therefore, scholars have questioned
spatial structures have far-reaching implications for
the application of monocentric models [13] and have
regional transport planning [6], risk assessment and
begun to apply polycentric regional density functions
policy decisions [5], public safety, and crisis manage-
to explain the spatial structural characteristics
ment [7]. Therefore, this study helps understand the
and evolutionary trends of regional population
characteristics of the regional spatial structure and its
density. Heikkila et al. [14] proposed three models of
evolutionary trends, thus providing a reference for
population density distribution in polycentric urban
regional spatial organization and restructuring.
Researchers have a long history of studying areas, followed by Small and Song [15], Bontjie [1],
regional spatial structures from the perspective of McMillen, McDonald [16] and Joseph [17] who applied
population spatial distributions. In 1951, Clark pio- polycentric regional density functions to Los
neered the Distance Decay Model, which suggests Angeles, Western Europe, Chicago, and Port-au-
that population density tends to decrease with Prince.
increasing distance from urban centers [8]. Since Research into population density models in
then, population density functions have been used China began in the late 1990s. This research focused
as effective tools to characterize urban spatial on fitting regional optimal population density
structures. On this basis, scholars such as Tanner, models based on census data and summarizing the
Sherratt, McDonald, Suits, Mason and others have characteristics of regional spatial structure evolution.
proposed various models of population density Zhou Chunshan and Xu Xueqiang were the first to
in monocentric cities [9], which have resulted in apply a monocentric city population density model to
significant achievements in the study of urban illustrate the urban population density development
population spatial structures. However, these studies process in Guangzhou [18]. Later, Wang and Zhou [19]
only depicted population distribution at the city and Luo and Wei [20] simulated the population density
level. Parr highlighted in the mid-1980s that the distribution in Beijing and Nanjing, respectively,
decaying characteristics of population density with under monocentric conditions. Along with rapid
distance in suburban and rural areas outside cities urbanization and industrialization in China, the
did not conform to a negative exponential model but spatial variation in population density has taken on
were closer to a Pareto function. He also proposed a new characteristics, and the regional spatial structure
unified functional form to describe the characteristics has gradually changed to a complex polycentric
of population density across regions [10]. Later, Bark- structure. Consequently, scholars have begun to
ley [11], Wang [4], and others applied different forms of explore the spatial structure of polycentric cities and
monocentric regional population density models to population spatial distribution. Wang and Meng [21]

62
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

used Shenyang as the study object and were the first analysis [28] and spatial autocorrelation analysis [29,30];
to investigate the polycentric population density however, few studies have used regional density
distribution structure in Chinese cities. Feng Jian and functions for analysis. Population density patterns
Zhou Zhouxing conducted a systematic study on the are important features of regional economies and so-
multi-core population density model in Beijing [22]. cieties [31]. The optimization of the spatial distribution
Wu Wenyu and Ma Xiya used Shanghai as the study of the population can have a significant impact on
object to analyze the characteristics of the polycentric the structure and layout of various factors, such as
structure of population distribution from 1990 to the economy, culture, and industry, helping to pro-
2000 [23] . Jiang Li and Wu Shoulong measured mote regional economic development and improve
the change in population spatial distribution in people’s quality of life. Viewing the growing impor-
Guangzhou [24], and the results demonstrated a good tance of the population spatial structure and rapid ur-
fit for the polycentric model. Subsequently, the banization and continuous economic restructuring in
analysis of polycentric structures gradually expanded Shandong Province in recent years, it is particularly
to regional areas, including the Beijing-Tianjin-He- essential to utilize the regional density function to
bei metropolitan area [25], Yangtze River Delta urban study its population distribution and spatial structure,
agglomeration [26] and Huanghuaihai Plain [27]. In evolution pattern and regional spatial structure’s
terms of research regions, most studies focused on variation trend.
more economically developed large cities or hotspot In summary, this study analyzed the population
regions. There has been less exploration of regions density characteristics of Shandong Province
than cities in terms of the volume of literature. In from 2000 to 2020 based on county data. It used
contrast to Western scholars’ systematic analysis the Inverse Distance Weighted method to clarify
of population density models and improved testing the spatial relationship of population distribution.
of model applications, the analysis of population Subsequently, the spatial structure of Shandong
density models in the Chinese geographical circle Province and its evolutionary characteristics were
primarily uses the existing population density model shown by applying monocentric and polycentric
to conduct empirical research on a specific region, regional density functions, respectively. The spatial
with few theoretical innovations. Moreover, Chinese agglomeration and diffusion of the population in
scholars mostly base their calculations on the Eu- Shandong Province were discussed in depth based
clidean distance. The Euclidean distance character- on the population growth rate and hot and cold spot
izes the straight-line distance, which differs from analysis. Analyzing the spatial structure of Shandong
the actual distance in reality and does not accurately Province from the standpoint of the regional density
reflect a real situation. When applying the regional function provides a new perspective on the study of
density function, the shortest vehicular route between spatial structure, which can be compared with the
two places, provided by Amap (https://www.amap. results obtained from other methods. Simultaneously,
com/), was adopted to describe the distance, which is the evolutionary characteristics and development
more relevant to a real situation than the Euclidean trend of Shandong Province’s population spatial
distance. structure can be examined, as well as the problems
In terms of population and economic output, that hamper its long-term development process and
Shandong is one of the largest provinces in the law of urban agglomeration development can be
China and is characterized by an uneven level of better understood. This can provide a basis for urban
development within the region. In previous studies, planners to rationally guide regional population
scholars have discussed the population distribution distribution, optimize resource allocation, make
in Shandong Province, starting from the structure in- scientific spatial planning decisions, and serve as a
dex [28,29], population gravity center [28,29], shift-share reference for other urban agglomerations in China.

63
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

2. Materials and methods Social Development of Shandong Province and

Vision 2035” that the institutional mechanism for the
2.1 Study area coordinated development of the whole region should
be promoted in an integrated manner. Optimizing and
Shandong Province is located in the econom- improving the regional spatial structure is the key
ically developed eastern coastal region of China, to achieving coordinated regional development [33],
downstream of the Yellow River, the second-longest which must be based on an in-depth understanding
river in China, and between the two state-level urban of the spatial structure of population density and its
agglomerations: the Beijing-Tianjin-Hebei region, evolutionary characteristics in Shandong Province.
and the Yangtze River Delta urban agglomerations.
To the east, it is separated from the Korean Peninsula
2.2 Methods
and Japan by the sea. Its eastern part is dominated
by hills, while the central part has many mountains, The regional density function is an important
and the southwest and northwest are mostly plains means to study the spatial structure of the regional
(Figure 1), with a total land area of 156,700 km2. population and its evolution trend. Since its
According to data from the 7th National Census, introduction, many scholars have extended its
the number of people living in Shandong Province application and supplemented its amendments,
exceeded 100 million, ranking second in China in achieving rich research results. Given that the
terms of total population. Simultaneously, the gross acquisition of population data is relatively
national product of Shandong Province is expected convenient and reliable and the mathematical form
to reach 728 million yuan in 2020, ranking third in of the regional density function is quite simple, this
China. Shandong Province is veritable in terms of its study also chose this method to model the population
population and economy. As an uncommon province agglomeration and diffusion characteristics of
in China where a province is developed as an urban Shandong Province.
agglomeration, the Shandong Peninsula Urban Ag- (i) Monocentric regional density function
glomeration is one of the important urban agglomer- The regional density function is used to describe
ations proposed in the National 14th Five-Year Plan how population density varies with distance from the
for development and growth and is also the core of central city, and is an effective tool for studying the
the national strategy to promote new urbanization characteristics of population agglomeration and dif-
and economic development [32] . To respond fusion [13]. Following Alperovich, the following four
positively to the national planning objectives, forms of monocentric area density functions (Table
Shandong Province specified in the “Outline of the 1) were validated by some scholars [4]:
14th Five-Year Plan for National Economic and The monocentric regional density function as-

Figure 1. Location and topography of Shandong Province.

64
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 1. Monocentric regional density functions and the parameters.

Function used Original
Models X variable Y variable Restrictions Meanings of parameters
in regression function
Linear Dr = a + br Same r Dr None r is the distance between the
Logarithmic Dr = a + blnr Same lnr r r≠0 area and the regional center,
D
Dr is the population density at
Power lnDr = a + blnr Dr = arb lnr lnDr r ≠ 0 and Dr ≠ 0 distance r from the regional
center, and a and b are
Exponential ln(Dr) = a + br Dr = aebr r lnDr Dr ≠ 0
parameters to be estimated.

sumes that there is only one central city in the re- and trend of regional population density under the
gion. Therefore, when discussing the form of the combined effect and influence of multiple centers.
monocentric regional density function, it is first nec- Heikkila et al. argued that if multiple centers are in
essary to divide Shandong Province into a number competition and division of labor with each other,
of monocentric metropolitan areas. In this study, we and there is both substitutability and complementari-
referred to Wang Fahui’s approach for delineating ty, then the polycentric density function may formal-
urban influence ranges [4] and used the urban areas of ly take the form of a merger of multiple monocentric
the central cities in the region as the core to delineate density functions. Based on this, they proposed three
their zones of influence. The delineation of the zone models for the distribution of population density in
of influence was based on the gravity model: polycentric urban areas [14]. Assumptions regarding
the impact of multiple centers on population distri-
Ii = Ci / rijβ  (1)
bution range from full substitution to full comple-
where Ci is the population size of city i; rij is the dis- mentation. Given the fact that most scholars deem
tance between regional unit j and city i; Iij indicates the impact of each center in the region to be between
the degree to which regional unit j is influenced or complete substitution and complementation [38] and
radiated by city i; and β is the distance friction coef- the actual situation in Shandong Province, this study
ficient, where the empirical value of β = 2.1 is taken selected the exponential model to test Shandong
from the gravity model used by Yang Qi in his study Province. Its functional form is:
of inter-regional passenger flows in China [34]. A cen- n
tral city is an attributed city of regional unit j if its Dr = å ai ebi ri (2)
radiation or influence on regional unit j is the great- i =1

est among all central cities. where Dr is the population density at distance r from
(ii) Polycentric regional density function the regional center, n represents the number of re-
The monocentric regional density function only gional centers, ri is the distance from the region to
considers the influence of a single central city when each center, and ai and bi are parameters specific to
analyzing the regional population distribution [25]. It regional center i.
is believed that the population density at the same (iii) Inverse Distance Weighted (IDW) Interpola-
distance from the center is the same, regardless of tion
the direction [17], which has some limitations when To clearly reflect the spatial relationship of the
analyzing the trend of population density changes population distribution in Shandong Province, this
in polycentric regions. Therefore, it is essential to study applied the IDW method to analyze the region-
consider the impact of multiple centers in the region al population density distribution. The IDW is one of
on the population distribution to more accurately in- the most commonly used methods for interpolating
terpret the characteristics of regional population ag- spatial information. It is based on the “first law of
glomeration and dispersion. The polycentric regional geography”, which assumes that the parameter val-
density function was used to reveal the distribution ues of observations closer to the prediction point are

65
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

more similar than those further away from the pre- the confidence intervals.
diction point [36]. Therefore, the IDW method assigns
more weight to observations that are closer to the 2.3 Data sources
predicted points [37]. The equation for this method is
as follows: In this study, resident population data for 2000,
2010, and 2020 were selected as the population data.
n Z n
Z = ∑ i / ∑dp Compared to the household population indicator, the
i (3)
i =1 d p i
i resident population indicator more accurately reflects
where Z is the estimated value, Zi is the i (i = 1,...,n) the total number of people residing in a region over a
sample value, di represents the distance, and p is the certain time period and can more accurately indicate
power of the distance, which was chosen based on the true size of a region’s population [41]. Resident pop-
the minimum mean absolute error criterion. The de- ulation figures for 2000, 2010, and 2020 were obtained
fault setting is p = 2, at which point the correspond- from the fifth, sixth, and seventh national census bulle-
ing method is called inverse distance squared inter- tins, respectively, published by each municipality.
polation [38]. For the administrative division data, the land area
(iv) Cold and hotspot analysis data of the county administrative area were used to
Cold and hotspot analysis is one of the methods calculate the regional population density, and the
used for global aggregation testing. It identifies high- data were derived from the Baidu Encyclopedia of
and low-value clusters of regional elements at differ- the corresponding county. The administrative divi-
ent spatial locations by calculating the Getis-Ord Gi* sions of Shandong Province underwent adjustments
statistics for each element in the dataset. That is, it and changes during the study period. To ensure con-
identifies the spatial distribution pattern of cold and tinuity and comparability of the data, this study used
hot spots in the study area [39]. The equation [40] is as the 2020 administrative divisions as the standard and
follows: matched the statistics for 2000 and 2010. This in-
n cluded correction and matching of the names used in
∑ Wij x j
Gi* =
j =1 2020 for county-level administrative units that had
n  (4)
∑ xj been renamed and adjusted, adjustment of the zoning
=j 1, j ≠ i of newly established districts and counties according
where Gi* is the Getis-Ord Gi* statistic. xj is the el- to the streets and towns under their jurisdiction in
emental attribute value at location j; n is the total 2020, and merging or splitting of county-level units
number of districts and counties; and W is the spatial with boundary adjustments and changes according
weighting matrix. When the Gi* value is positive and to boundary changes. In addition, non-administrative
significant, it indicates that the values around the areas, such as high-tech zones and economic and
detected points are relatively high and belong to the technological development zones in each municipal-
high-value spatial agglomeration (hot spot area); a ity were consolidated according to the area in which
negative and significant Gi* value indicates that the they were located.
values around the detected points are relatively low Data used to measure the distance between the
and belong to the low-value spatial agglomeration two regions were obtained from Amap (https://www.
(cold spot area); a Gi* value of zero indicates that the amap.com/). This study selected the shortest vehicle
result was randomly generated and not statistically route between the county or municipal government
significant. A significance test was performed on the of the two administrative regions to represent the
obtained Gi* values to identify cold and hot spots at distance between each region.

66
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

3. Results and discussions

3.1 Population distribution characteristics of
Shandong Province

Statistical characteristics of the population dis-

tribution
An analysis of regional population density can re-
veal the basic characteristics of regional population
distribution and reflect regional differences in pop-
ulation distribution [27]. According to the population
density distribution map of Shandong Province from
2000 to 2020 (Figure 2), the population of Shan-
dong Province was spatially unevenly distributed,
with obvious differences between districts and coun-
ties. Areas with high population densities were more
dispersed, with multiple densely populated circles
in local areas. In contrast, areas with low population
densities were characterized by faceted contiguity.
Specifically, the population density was lower in
the northern part of Shandong Province and higher
in the eastern coastal, central, and southern regions.
Peak population density areas were concentrated in
the urban areas of the cities, with the northern and
southern districts of Qingdao having extremely high
population densities, forming two polarized centers
with population densities of over 10,000 persons per
square kilometer in all years. The other regions with
higher population densities (Lixia District in Jinan,
Zhangdian District in Zibo, Licang District in Qing-
dao and Zhifu District in Yantai) were almost all
between 5,000 and 10,000 persons per square kilo-
meter, a far cry from the two districts in Qingdao.
There were sharp differences in population density
between urban areas and districts and counties out- Figure 2. Population density distribution map of Shandong
side urban areas, presenting a clear decay from urban Province.
areas to the periphery of the city. Most population
densities in the districts and counties outside the city square kilometer, a difference of nearly 150 times.
are less than 1,000 persons per square kilometer. In Specifically, in terms of spatial changes in popula-
addition, the population density varied significantly tion distribution, with the exception of Heze City and
between districts and counties in Shandong Province the peripheral districts and counties of Qingdao City,
(Table 2). For example, in 2000, the minimum pop- where there was a widespread increase in population
ulation density was only 97 persons/square kilome- density, there was little change in the distribution of
ter, whereas the maximum reached 14,918 persons/ all classes of population density over the study pe-

67
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Table 2. The descriptive statistics of population density in Shandong Province.

Population density (person/km2)

Year
Maximum Minimum Average Standard deviation
2000 14918 97 1010 1889
2010 18162 109 1168 2246
2020 17411 108 1280 2337
2
Number of counties by population density(person/km )
Year
< 300 300-500 500-700 700-1000 1000-3000 3000-5000 > 5000
2000 7 38 52 15 18 2 4
2010 7 36 45 19 20 5 4
2020 8 41 34 20 21 6 6

riod, and the pattern of population distribution was higher than that of other districts and counties. The
relatively stable. influence of the natural geographical environment
From the perspective of temporal changes in pop- on population distribution can be observed in the to-
ulation density, both the mean and standard deviation pography of the terrain. The flatter terrain favors the
of population density in Shandong Province showed growth of local crops, which in turn leads to higher
an increasing trend (Table 2). This indicates that the population densities [42]. In the south-central region
population density of Shandong Province has grown of Shandong Province, mountainous and hilly terrain
over the past 20 years, and regional differences have is widespread, transportation is inconvenient, and
increased. This is demonstrated by the minimum economic development is relatively slow, making
population density in Shandong Province (Hekou the population more dispersed [43]. The Yellow River
District, Dongying City) increasing from 97 persons/ Delta region of Dongying and Binzhou, although
km2 in 2000 to 108 persons/km2 in 2020; the average located on the plain, is plagued by seawater erosion
population density increasing from 1010 to 1280 per- and serious land salinization, making reclamation
sons per square kilometer over a period of 20 years; and renovation difficult; therefore, it is sparsely
and the number of areas with a population density populated. The topography of the Jiaodong Penin-
greater than 1,000 persons per square kilometer ris- sula is predominantly hilly and complex; thus, the
ing from 24 in 2000 to 33 (Table 2), with new areas population is mostly concentrated in gently sloping
including Licheng District in Jinan, Laoshan District coastal areas [28]. Coupled with the advantageous ge-
in Qingdao, Laishan District in Yantai, Donggang ographical location of the Shandong Peninsula urban
District in Rizhao, Luozhuang District, and Hedong agglomeration, convenient waterway transportation,
District in Linyi. There was a marked change in the earlier development, and higher levels of economic
districts and counties with population density values development and urbanization than in inland areas,
between 500 and 700 persons/km2, with several of this has led to the movement of the surrounding
them having increased to 700-1000 persons/km2 in population to coastal areas, resulting in significant
2020. differences between the surrounding districts and
In summary, the densely populated areas of Shan- counties [44]. In addition, the areas along the axes
dong Province are concentrated in economically of the main railway lines (Jiaoji, Beijing-Shanghai,
developed areas with superior natural geography. Beijing-Kowloon and Xinshi lines) are also areas of
Economically developed areas with accelerating lev- high population density.
els of urbanization are attracting growing population
concentrations from outlying districts and counties, Spatial characteristics of population distribution
making the local population density significantly The IDW results (Figure 3) demonstrated that,

68
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

high-value areas tended to diffuse from the center to

the periphery, presenting a certain circular structure.
Moreover, degradation of the high- and low-value
areas occurred in almost all outer districts of the city
during the study period. This could indicate a devel-
opmental tendency of the population to concentrate
on the central city.
Overall, the results based on the IDW revealed
the stable existence of six statistically significant
high population density areas within Shandong Prov-
ince, indicating a polycentric spatial structure char-
acteristic of regional population distribution. The
six densely populated areas centered on the Jinan,
Qingdao, Zibo, Weifang, Yantai, and Dezhou centers
can serve as the six population density centers in
Shandong Province. This finding provided a basis
for the application of the polycentric regional densi-
ty function to analyze the characteristics of regional
population agglomeration and diffusion.

3.2 Agglomeration and diffusion of the

population in Shandong Province

Analysis based on the monocentric regional

density functions
According to the results of the IDW, there are six
statistically significant population density centers in
Shandong province, namely Jinan center (Lixia Dis-
trict, Shizhong District, Huaiyin District, Tianqiao
District, Licheng District), Qingdao center (Shinnan
District, Hebei District, Laoshan District, Licang
District), Zibo center (Zichuan District, Zhangdian
District, Zhoucun District), Yantai center (Zhifu
Figure 3. Population density distribution map of Shandong District, Fushan District, Muping District, Leshan
Province based on IDW.
District), Weifang center (Weicheng District, Hant-
in general, the areas with high and low values did ing District, Fangzi District, Kuiwen District) and
not vary significantly during the study period. The Dezhou center (Decheng District). Considering the
divergence in population density was evident, show- need for computation, this study chose the shortest
ing a trend of local concentration in high-value areas route between the other district and county govern-
and a contiguous distribution in low-value areas. ments and the regional center government, plus one
A spatial distribution pattern was formed, with Ji- kilometer as the distance from the regional center;
nan, Qingdao, Zibo, Weifang, Yantai, and Dezhou and thus, the impact area was divided. Combined
as local high-value areas, whereas other areas were with the results of the influence zone division, this
low-value areas for a long time. Meanwhile, several study fits four forms of density functions for the Ji-

69
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

nan, Qingdao, Zibo, Yantai, Weifang, and Dezhou inent from 2010 to 2020, while the growth in the
centers. The results demonstrated (Table 3) that the remaining four population density centers was rela-
logarithmic model exhibited high goodness of fit in tively stable over the two decades, with the Qingdao
all six regions. Therefore, a logarithmic model was and Dezhou centers rising particularly rapidly. An
selected for analysis of the study area. increase in the intercept reflects an increase in pop-
The intercept a and slope b of the regional density ulation density in the urban areas of the central city,
function represent the estimated population density at while an increase in the slope indicates that popula-
the center of the region and the regional population tion density decreases more rapidly as the distance to
density gradient, respectively. By observing the fitted the central city increases. This indicates a clear and
lines of density functions based on the logarithmic increasing trend of centripetal agglomeration in the
model for the Jinan, Qingdao, Zibo, Yantai, Weifang, six regions over the 20-year period. Districts closer
and Dezhou centers over the study period (Figure 4), to the center not only have a higher population densi-
the absolute values of the intercept a and slope b in- ty but also a faster growth rate in population density
creased in all six regions from 2000 to 2020, and the than more distant districts. After analyzing the cen-
growth in the Jinan and Dezhou centers was prom- sus data, this study suggests that the rapid increase
Table 3. Regression results of monocentric regional density functions.

Function a b R2
Region
form 2000 2010 2020 2000 2010 2020 2000 2010 2020
* * *
Linear 786.652 919.785 1085.737 –2.793 –3.767 –5.605 0.226 0.235 0.238
Logarithmic 1310.016 1624.713 2145.855 –187.243 –251.158 –372.169 0.801 0.811 0.822
Jinan
Power 7.123 7.321 7.556 –0.216 –0.248 –0.309 0.687 0.694 0.685
* * *
Exponential 6.517 6.619 6.676 –0.003 –0.004 –0.005 0.190 0.193 0.179
* * * * * *
Linear 1860.429 2478.311 3037.079 –14.989 –24.340 –30.040 0.409 0.443 0.502
Logarithmic 3163.444 4045.571 4738.781 –617.727 –834.934 –958.835** 0.921 0.945 0.929
Qingdao ** *
Power 8.053 8.353 8.577 –0.409 –0.458 –0.471 0.920 0.901 0.837
* * *
Exponential 7.261 7.598 7.900 –0.011 –0.016 –0.018 0.511 0.575 0.683
Linear 1203.035 1356.848 1469.173 –13.461 –16.179 –18.770* 0.873 0.895 0.850
**
Logarithmic 1328.238 1533.722 1730.581 –194.064 –241.313 –297.126 0.780 0.856 0.917
Zibo ** ** **
Power 7.264 7.419 7.546 –0.242 –0.271 –0.310 0.655 0.718 0.768
Exponential 7.150 7.275 7.343 –0.018 –0.020 –0.021 0.830 0.874 0.854
** ** * * * *
Linear 567.085 716.164 841.868 –1.030 –2.778 –4.409 0.059 0.238 0.308
** ** ** * * *
Logarithmic 629.576 813.582 988.158 –36.753 –81.059 –126.524 0.155 0.421 0.528
Yantai
Power 6.507 6.686 6.860 –0.075* –0.139* –0.200* 0.167 0.364 0.397
* * *
Exponential 6.317 6.507 6.581 –0.002 –0.005 –0.006 0.066 0.192 0.193
** * *
Linear 562.077 621.762 689.244 –1.242 –0.937 –2.967 0.283 0.041 0.307
Logarithmic 560.444 742.276 924.003 –16.906* –47.690* –110.551** 0.115 0.184 0.735
Weifang * * **
Power 6.330 6.593 6.807 –0.035 –0.076 –0.162 0.130 0.174 0.603
* * *
Exponential 6.328 6.398 6.463 –0.002 –0.001 –0.004 0.294 0.038 0.251
Linear 2358.959* 2904.270* 4286.715* –52.823* –67.412* –106.679* 0.937 0.942 0.939
** ** ** ** ** **
Logarithmic 2383.428 2932.550 4334.992 –542.267 –690.924 –1094.709 0.993 0.994 0.993
Dezhou ** ** **
Power 7.771 7.979 8.365 –0.468 –0.515 –0.648 0.986 0.988 0.985
Exponential 7.745 7.956 8.325** –0.045* –0.050* –0.063* 0.924 0.933 0.919
Note: The significance level of parameter estimates without labeling is p = 0.001; * indicates that the parameter estimates are not significant, and ** indicates that the sig-
nificance level of parameter estimates is p = 0.01.

70
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

in population density in the centers of Jinan and it is developing well. Weifang ranks among the top
Qingdao is mainly due to the large influx of people cities in Shandong Province in terms of geographical
from other areas. The population of the urban area in area, population, and total economic scale, and has a
Jinan is large, and the growth of its population densi- strong industrial base as a traditional industrial and
ty is closely related to its special status as the capital agricultural city. It relies on the Weifang Port to ac-
city of the province. The population density in the tively develop its port economy. In recent years, Wei-
Jinan center rose faster from 2010 to 2020, and this fang has been optimizing and upgrading its industrial
study speculates on the following reasons: In 2017, system, improving its modern industrial system, and
Shandong Province clarified the strategic concept of continuously increasing its investment in innovative
the conversion of old and new kinetic energy, and cities with certain advantages. The transport location
Jinan city actively responded to the requirements and natural conditions provide a good foundation for
by developing a number of new industries. The ob- the development of Dezhou. Relying on its superior
jective demand for economic structure optimization transportation location, the Dezhou center has taken
and policy support led to the migration of talent to the lead in development, driven by the synergistic
Jinan. In addition, in 2020, Jinan proposed a “strong development of the Beijing-Tianjin-Hebei Region.
provincial capital” strategy and zero-threshold settle- As a large national transportation hub, Dezhou is
ment policy, which also contributed to rapid growth developing at a relatively rapid pace. In addition,
in population density. Qingdao, the most developed rich mineral resources exist underground in Decheng
leading city in the Jiaodong Economic Circle and an District. The Dezhou Power Plant of Huaneng Inter-
important national coastal center city, is well placed national Power Co., Ltd., the second largest thermal
in terms of economic output, employment opportu- power plant in China, is located in the Decheng Dis-
nities, social security, and infrastructure. Simulta- trict, which also attracts the population of surround-
neously, Qingdao hosted various major events, such ing districts and counties to work there [45]. Although
as the Shanghai Cooperation Organization summit, the Zibo, Yantai, Weifang, and Dezhou centers also
and the city’s influence increased significantly. The have a certain floating population, they are inferior
reform of the household registration system and the to the Jinan and Qingdao centers.
relaxation of settlement and access conditions also Furthermore, it can be observed from Figure 4
provided conditions for population agglomeration, that the growth rate varied considerably between the
which undoubtedly attracted the population of the different areas in the hinterlands of the Qingdao and
surrounding districts and counties. As far as the other Dezhou centers. The growth in population density
three centers are concerned, Zibo has a long history was very significant in proximity to the Qingdao
of industry and is an old industrial area that has de- and Dezhou centers, whereas the growth in areas
veloped based on its resources. This, coupled with its further away from them was slow. This is in line
emphasis on the transformation and upgrading of tra- with the “core growth-hinterland stagnation” pattern
ditional industries and its focus on the tertiary sector of Barkley et al. [11]. In contrast, the growth patterns
during its development in recent years, has allowed of the remaining centers were more moderate, and
the region to flourish. Yantai is strategically located although the areas closer to the center continued to
and is an important land and sea transportation hub grow faster than those away from it, the difference in
in Northeast Asia and the Circum-Bohai Sea Eco- growth rates was more moderate than that in the two
nomic Zone. Relying on its excellent port conditions, centers mentioned above. These two growth patterns
the city is actively cultivating and growing its hub reflect the difference between the development of
economy. Simultaneously, as one of the “three cores” the central city and that of its hinterland. Qingdao’s
of Shandong Province’s comprehensive pilot zone economy is large and has an extremely strong ag-
for the conversion of old and new kinetic energy, glomeration effect. It has a large development gap

71
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 4. Growth patterns in different population density centers.

with the surrounding districts and counties, and the of economic development in Dezhou as a whole.
city of Qingdao does not have a strong radiation Compared to other county-level regions, Decheng
drive. The lack of large-scale, complete industrial District has a high degree of development. In De-
chains and high-level modern industrial enterprises zhou city, the center has a comparative advantage. In
in the Jiaodong Peninsula makes it difficult to stimu- addition, most new infrastructure is concentrated in
late economic development in the hinterland [46]. The cities and counties, which increases the development
stronger polarization effect led to a larger difference gap [47]. The center of Jinan has a well-developed
in population density between the central area and economy, excellent service facilities and, thanks to
peripheral districts and counties. The reason for the the integration construction of the Jinan metropolitan
large difference in population density between the area, Jinan and its neighboring cities have reached
center of Dezhou and the periphery is the low level the standard of a “one-hour economic circle” with

72
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

convenient transportation [48]. The diffusion effect has Qingdao, their development is inevitably influenced
also led to higher population densities in neighboring by the two powerhouses, limiting their polarization
areas, with stronger spatial correlations between re- and diffusion effects.
gional units [42]. In addition, Figure 4 shows that the
Analysis based on the polycentric regional den-
population density in Jinan center is lower than that
sity functions
in Qingdao center. This study speculates that this
is because of the rapid development of Qingdao in The determination of regional density centers was
recent years, with Jinan, the provincial capital city, based on the above analysis of the characteristics
being overwhelmed by Qingdao, and the increasing of the population density distribution in Shandong
competition gap. Compared to Jinan, the gap be- Province. Six population density centers were used
tween the centers of Zibo, Yantai, and Weifang and for the monocentric regional density function anal-
the peripheral areas is smaller. The common reason ysis: Jinan, Qingdao, Zibo, Yantai, Weifang, and
for this is that the economic development of the dis- Dezhou. According to the IDW, the most prominent
tricts and counties within the sphere of influence of centers were Jinan, Qingdao, and Zibo. The polycen-
these three centers is at a similar level. The centers tric regional density function followed the six
do not have a strong population concentration capac- centers mentioned above, and the models were fitted
ity, but their relatively rapid economic development and compared separately for different numbers of
has attracted a certain population from the surround- centers. The fit results for the polycentric population
ing districts and counties. Together with the prox- density model are presented in Table 4.
imity of these three centers to the cities of Jinan and The polycentric population density model based

Table 4. Comparison of the fitting results of the polycentric population density model.
Model 1 Model 2 Model 3 Model 4
Center Year
a b a b a b a b
2000 641.290 –0.002 1538.991 –0.482 1359.701 –0.356 1319.043 –0.324
Jinan 2010 1020.603 –0.007 1459.236 –0.114 1467.612 –0.124 1444.702 –0.100
2020 3223.348 –0.568 2154.727 –0.111 2137.374 –0.094 2138.823 –0.096
2000 3216.818 –0.068 3138.012 –0.080 3144.400 –0.080 3144.892 –0.080
Qingdao 2010 4267.686 –0.046 3932.811 –0.065 3918.105 –0.066 3928.841 –0.066
2020 4789.047 –0.044 4551.754 –0.051 4570.155 –0.051 4568.794 –0.051
2000 775.341 –0.050 855.579 –0.044 846.112 –0.045 846.598 –0.045
Zibo 2010 1016.212 –0.040 1053.177 –0.045 1044.845 –0.047 1024.394 –0.049
2020 998.269 –0.006 1284.258 –0.051 1267.076 –0.052 1267.173 –0.052
2000 498.867 0.000 241.642 –0.272 234.610 –0.236
Yantai 2010 504.216 0.000 528.988 0.000 486.983 –0.268
2020 504.531 0.000 805.472 –0.309 771.650 –0.273
2000 504.355 0.000 504.311 0.000
Weifang 2010 486.365 –0.547 568.621 –0.001
2020 538.579 –0.001 534.415 –0.001
2000 3972.119 –0.716
Dezhou 2010 2940.236 -0.162
2020 7021.046 -0.577
2000 0.887 0.958 0.958 0.968
2
R 2010 0.888 0.955 0.946 0.962
2020 0.920 0.937 0.940 0.961

73
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

on the exponential form fits better and is generally ward urban areas. In terms of changes in b value of
more applicable than the monocentric model, with the population density gradient, both the Jinan and
R2 values above 0.9 in all years of the study period. Qingdao centers exhibited a decrease in the b value
This indicates that the exponential model can better during the study period, demonstrating continuous
explain the structure of the population density dis- outward population diffusion. This reflected the
tribution in Shandong Province, which has formed a spatially evolving characteristics of the population
polycentric population density spatial structure, and structure of the center’s growth-based diffusion pat-
the relationship between population density centers tern, where the center’s population density increased
is between complete substitution and complemen- while the population density gradient decreased.
tation. Model 1 used three centers, Jinan, Qingdao, This indicated that Jinan and Qingdao grew stronger
and Zibo, all of which had R2 values exceeding 0.88, as centers while driving the development of other
indicating that these three centers explained, to a districts and counties within their spheres of influ-
large extent, the distribution of population density in ence, with a high degree of center development. The
Shandong Province. By 2020, the R2 value reached Zibo, Yantai, and Weifang centers displayed a gener-
0.92, indicating that these three centers had an in- ally weak upward trend, with the population density
creasingly strong influence on regional population gradient increasing but not changing significantly.
density distribution and could dominate regional This reflected the centripetal concentration pattern of
population density distribution. Model 2 added the the population and the fact that the three areas were
Yantai center, and the R2 value further increased to relatively “weak” centers. The population density
over 0.93. The explanatory power of the model was gradient in the Dezhou center demonstrated a decline
further enhanced, indicating that these four centers followed by an increase, indicating that the Dezhou
could fully explain the regional population densi- center had not yet shown a strong agglomeration
ty distribution. However, between 2000 and 2020, ability from 2000 to 2010. The regional population
the R2 values showed a decreasing trend, indicating distribution was mainly dominated by centers such
as Jinan, Qingdao, and Zibo, while the agglomera-
that other population density centers in the region
tion effect gradually emerged after 2010, displaying
continued to develop and had an increasingly sig-
a trend of population agglomeration toward the
nificant impact on the regional population density
center.
distribution, while the explanatory power of the Yan-
tai center on the overall regional population density
distribution decreased. Models 3 and 4 added the
3.3 Further discussion on population agglom-
eration and diffusion
Weifang and Dezhou centers, respectively, and the
results demonstrated that the two centers also had To further explore the spatial variation pattern
an impact on the regional population density distri- of population dispersion in Shandong Province, we
bution; however, there was no significant change in first visualized the population growth rate during the
R2 values, indicating that the Weifang and Dezhou study period using ArcGIS, and the results are shown
centers made a relatively limited contribution to the in Figure 5. Overall, there were significantly more
degree of model interpretation. areas with positive population growth rates from
Changes in the estimated values of a and b reflect 2000 to 2010 than from 2010 to 2020, and the areas
the characteristics of regional population agglomer- with higher population growth rates were broadly
ation and diffusion at different stages. The estimates concentrated in the central districts and counties of
of regional population density a showed that all the cities. Specifically, from 2000 to 2010, the popu-
six centers had an upward trend in their population lation growth rate in the western region of Shandong
density estimates. This reflected the overall trend of Province was higher than that in the eastern region,
population concentration in Shandong Province to- with a contiguous positive regional distribution of

74
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Figure 5. Population growth rate map and distribution map of hot and cold spots in Shandong Province.

growth rates. However, the eastern regions of Ping- above 4% in the period 2010-2020.
du City, Kuiwen District, Fangzi District, Laoshan Based on the characteristics of the population
District, Huangdao District, Laishan District and growth rate of each county and district in Shandong
Ju County had the highest population growth rates. Province, this study further explored the agglomera-
The regions with the lowest population growth rates, tion and diffusion characteristics using cold and hot
namely Jimo, Chengyang, Licang, and Lanshan, spot analyses. As shown in Figure 5c, the spatial
were also located in the eastern part of Shandong variations in the cold and hot spot areas from 2000
Province. This indicated that the population growth to 2010 were small, with more hot spot areas than
rate varied more among districts and counties in the cold spot areas. Specifically, there are six hotspot
eastern region and less among districts and counties areas, mainly in the eastern part of Shandong Prov-
in the western region. For the period 2010-2020, ince, including Pingdu City, Fangzi District, Laos-
regions with negative population growth rates con- han District, Huangdao District, Leshan District,
stituted the majority. The population growth rate be- and Juxian County. There are four cold spots: Jimo
tween city center districts and peripheral districts in- District, Chengyang District, Licang District, and
creased further. This reflected a further strengthening Lanshan District, which are more concentrated than
of population agglomeration and a more pronounced the hot spots. Figure 5d shows that there was some
polarization effect of the centers. It is worth noting replacement of cold and hot spot zones in 2010-2020
that most of the regions with the highest growth rates compared with those in 2000-2010. Four cold spot
in the period 2000-2010 had growth rates of less zones—Pingdu City, Laoshan District, Huangdao
than –4% in the period 2010-2020. In contrast, the District, and Ju County—were hot spot zones during
regions with the lowest growth rates in the period the last time period. There are five hotspot zones: the
2000-2010 almost all saw their growth rates increase Zhoucun, Lanshan, Jimo, Chengyang, and Licang

75
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Districts. Except for Zhoucun District, the other four land differed. Conversely, the polycentric regional
districts were cold spots during the last time period. density function reflected that some of the centers
In summary, the population agglomeration effect displayed a spatial trend of diffusing outward, and
in the central districts and counties of Shandong that different levels of centers exhibited different
Province was increasing. Variations in population patterns of population agglomeration and diffusion.
size were more significant along the eastern coastal The spatial statistical analysis of population densi-
area than in the inland area, and the phase of popula- ty growth further validated the polycentric density
tion growth rates along the eastern coastal area was function. The multicenter regional density function
more evident over the study period. reflected the trend of population density change
more accurately and precisely because of the com-
4. Conclusions and limitations prehensive consideration of the influence of multiple
centers on the distribution of population density.
4.1 Conclusions
4.2 Limitations
This study analyzed the distribution of population
density in counties in Shandong Province based on The study of the characteristics of regional popu-
data from the 2000, 2010, and 2020 censuses. To lation density distribution is important for identifying
directly reflect the spatial relationship of the pop- the characteristics of regional spatial structure devel-
ulation distribution, this study used the IDW. On opment. In this study, monocentric and polycentric
this basis, the monocentric and polycentric regional regional density functions were discussed, and the
density functions were applied to study the charac- analysis and comparison were relatively complete.
teristics of population agglomeration and diffusion in However, this study only started with demograph-
Shandong Province, and the changes in the popula- ic data, as a more diverse range of socioeconomic
tion growth rate and variations in cold and hot spots data was required for a comprehensive analysis of
were further analyzed over a twenty-year period. The the regional spatial structure. Simultaneously, the
results indicated that the spatial distribution of the distribution of the population density is influenced
population in Shandong Province was uneven, with by economic, social, and natural factors, and further
obvious differences in population size and changes exploration of the mechanisms that influence the
between districts and counties. With urban areas as spatial structure of the population is needed. In ad-
the center of population concentration, a multi-center dition, this study was based on county data and did
spatial structure was formed, and six central cities not discuss the spatial structure of the population of
dominated the spatial distribution trend of regional Shandong Province in sufficient detail; a more com-
population density, including Jinan, Qingdao, Zibo, prehensive study could be conducted in the future
Yantai, Weifang, and Dezhou centers. Methodologi- using data from smaller-scale geographical units.
cal and conclusive differences between monocentric
and polycentric regional density functions were com-
Author Contributions
pared. The polycentric regional density function ex-
hibited a better fit and was more applicable than the Conceptualization (Chen YB). Methodology
monocentric model. The results of the monocentric (Chen YB, Zhao XH). Software (Chen YB, Zhao
regional density function analysis showed a clear and XH). Formal analysis (Zhao XH). Resources (Zhao
increasing trend toward the centripetal agglomera- XH). Data curation (Zhao XH). Writing of the origi-
tion of regional centers over the study period, but the nal draft (Zhao XH). Writing, reviewing, and editing
ability to agglomerate and the difference in the rate (Chen YB). Visualization (Zhao XH). Supervision
of population growth between the center and hinter- (Chen YB).

76
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

Conflict of Interest [7] Li, J.G., Li, J.W., Yuan, Y.Z., et al., 2019. Spa-
tiotemporal distribution characteristics and
There is no conflict of interest. mechanism analysis of urban population densi-
ty: A case of Xi’an, Shaanxi, China. Cities. 86,
Funding 62-70.
This research was funded by the Shandong Pro- DOI: https://doi.org/10.1016/j.cities.2018.12.008
vincial Natural Science Foundation (grant number [8] Clark, C., 1951. Urban population densities.
ZR202102240088). Journal of Royal Statistics Society. 114(4), 490-
496.
DOI: https://doi.org/10.2307/2981088
References [9] Cheng, L., 2014. Zhong guo qu yu ren kou mi
[1] Bontje, M., 2001. Dealing with deconcentration: du mo xing ji jing ji zeng zhang de kong jian mo
population deconcentration and planning re- shi (Chinese) [Regional population density func-
sponse in polynucleated urban regions in North- tion and spatial patterns of economic growth in
west Europe. Urban Studies. 38(4), 769-785. China] [Ph.D.’s thesis]. Changchun: Northeast
DOI: https://doi.org/10.1080/00420980120035330 Normal University. p. 8-9.
[2] Salvati, L., Carlucci, M., 2016. In-between sta- [10] Parr, J.B., 1985. A population—density ap-
bility and subtle changes: Urban growth, popu- proach to regional spatial structure. Urban Stud-
lation structure, and the city life cycle in Rome. ies. 22(4), 289-303.
Population Space and Place. 22(3), 216-227. DOI: https://doi.org/10.1080/00420988520080531
DOI: https://doi.org/10.1002/psp.1877 [11] Barkley, D.L., Henry, M.S., Bao, S., 1996. Iden-
[3] Kabisch, N., Haase, D., Haase, A., 2012. Urban tifying “Spread” versus “Backwash” effects in
Population development in Europe, 1991-2008: regional economic areas: A density functions
The examples of Poland and the UK. Interna- approach. Land Economics. (72), 336-357.
tional Journal of Urban and Regional Research. [12] Limtanakool, N., Dijst, M., Schwanen, T., 2007.
36(6), 1326-1348. A theoretical framework and methodology for
DOI: https://doi.org/10.1111/j.1468-2427.2012.01114.x characterising national urban systems on the
[4] Wang, F.H., 2001. Regional density functions basis of flows of people: Empirical evidence for
and growth patterns in major plains of China, France and Germany. Urban Studies. 44(11),
1982-1990. Papers in Regional Science. 80(2), 2123-2145.
231-240. DOI: https://doi.org/10.1080/00420980701518990
DOI: https://doi.org/10.1007/PL00013624 [13] Subasinghe, S., Wang, R., Murayama, Y., 2022.
[5] Jia, P., Qiu, Y.L., Gaughan, A.E., 2014. A fine- Modelling spatial pattern of population distribu-
scale spatial population distribution on the tion in 50 largest cities in the world: A geospa-
high-resolution gridded population surface and tial approach. IOP Conference Series: Earth and
application in Alachua County, Florida. Applied Environmental Science. 1109(1), 012065.
Geography. 50, 99-107. DOI: https://doi.org/10.1088/1755-1315/1109/1/012065
DOI: https://doi.org/10.1016/j.apgeog.2014.02.009 [14] Heikkila, E., Gordon, P., Kim, J., et al., 1989.
[6] Murray, A.T., Davis, R., Stimson, R.J., et al., What happened to the CBD—distance gradient?
1998. Public transportation access. Transporta- Land values in a polycentric city. Environment
tion Research Part D: Transport and Environ- and Planningn A. 21(2), 221-232.
ment. 3(5), 319-328. [15] Small, K.A., Song, S., 1994. Population and
DOI: https://doi.org/10.1016/S1361-9209(98) employment densities: Structure and change.
00010-8 Journal of Urban Economics. 36, 292-313.

77
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

DOI: https://doi.org/10.1006/juec.1994.1037 2006.12.008

[16] McMillen, D.P., McDonald, J.F., 1998. Subur- [24] Jiang, L., Wu F.L., 2013. Guang zhou ren kou
ban subcenters and employment density in met- kong jian fen bu bian dong yu duo zhong xin
ropolitan Chicago. Journal of Urban Economics. cheng shi kong jian jie gou yan hua ce du (Chi-
43(2), 157-180. nese) [Guangzhou population spatial distribution
DOI: https://doi.org/10.1006/juec.1997.2038 and polycentricity spatial structure evolution].
[17] Joseph, M., Wang, F.H., 2010. Population densi- Tropical Geography. 33(2), 147-155.
ty patterns in Port-au-Prince, Haiti: A model of DOI: https://doi.org/10.13284/j.cnki.rddl.002318
Latin American city? Cities. 27(3), 127-136. [25] Sun, T.Sh., Li, G.P., Lu, M.H., 2009. Jing jin ji
DOI: https://doi.org/10.1016/j.cities.2009.12.002 dou shi quan ren kou ji ju yu kuo san ji qi ying
[18] Zhou, Ch.S., Xu, X.Q., 1997. Guang zhou shi xiang yin su: Ji yu qu yu mi du han shu de shi
ren kou kong jian fen bu te zheng ji yan bian qu zheng yan jiu (Chinese) [Concentration and de-
shi fen xi (Chinese) [Population distribution and centralization of population in the Beijing-Tian-
its change in Guangzhou city]. Tropical Geogra- jin-Hebei metropolitan region and its determi-
phy. 17(1), 53-60. nants: A regional density function approach].
[19] Wang, F.H., Zhou, Y.X., 1999. Modelling urban Acta Geographica Sinica. 64(8), 956-966.
population densities in Beijing 1982-90: Subur- DOI: https://doi.org/10.11821/xb200908007
banisation and its causes.Urban Studies. 36(2), [26] Liu, N.Q., Deng, M., 2018. Duo zhong xin jie
271-287. gou mo shi yu zhang san jiao cheng shi qun ren
DOI: https://doi.org/10.1080/0042098993600 kou kong jian fen bu you hua (Chinese) [Urban
[20] Luo, J., Wei, Y.D., 2006. Population distribu- spatial structure and optimization of population
tion and spatial structure in transitional Chinese distribution in Yangtze River Delta Region]. Re-
cities: A study of Nanjing. Eurasian Geography view of Industrial Economics. 27(4), 91-103.
and Economics. 47(5), 585-603. DOI: https://doi.org/10.19313/j.cnki.cn10-1223/
DOI: https://doi.org/10.2747/1538-7216.47.5.585 f.2018.04.007
[21] Wang, F.H., Meng, Y., 1999. Analyzing urban [27] Hong, Sh.M., Hao, J.M., 2018. Huang huai hai
population change patterns in Shenyang, China ping yuan ren kou kong jian jie gou ji qi yan
1982-90: Density function and spatial associa- bian te zheng yan jiu (Chinese) [Spatial structure
tion approaches. Geographic Information Sci- of population and characteristics of evolution in
ences. 5(2), 121-130. Huang-Huai-Hai Plain]. Resource Development
DOI: https://doi.org/10.1080/10824009909480521 & Market. 34(11), 1526-1532+1557.
[22] Feng, J., Zhou, Y.X., 2003. Jin 20 nian lai bei DOI: https://doi.org/10.3969/j.issn.1005-8141.
jing dou shi qu ren kou zeng zhang yu fen bu 2018.11.008
(Chinese) [The growth and distribution of pop- [28] Dong, L.F., Wang, L.L., Shao, J.H., et al., 2012.
ulation in Beijing Metropolitan Area (1982- Jin 20 nian lai shan dong sheng ren kou fen bu
2000)]. Journal of Geographical Sciences. 58(6), kong jian ge ju ji qi yan bian te zheng fen xi
903-916. (Chinese) [Population spatial distribution pat-
DOI: https://doi.org/10.11821/xb200306014 tern and evolution characteristics of Shandong
[23] Wu, W.Y., Ma, X.Y., 2006. Duo zhong xin province in recent 20 years]. Territory & Natural
cheng shi ren kou mo xing ji mo ni: Yi shang hai Resources Study. 138(3), 11-13.
wei li (Chinese) [Polycentric population density DOI: https://doi.org/10.16202/j.cnki.tnrs.2012.03.031
functions and modelling: The case of Shanghai]. [29] Li, W.X., Sun, X.H., 2007. Ji yu GIS de shan
Modern Urban Research. 21(12), 39-44. dong sheng ren kou zhong xin qian yi yan jiu
DOI: https://doi.org/10.3969/j.issn.1009-6000. (Chinese) [Study of the migration of population

78
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

gravity center in Shandong Province based on [37] Ghorbani, Z., Sexton, A., Van Loon, L.L., et al.,
GIS]. Journal of Shandong Normal University. 2023. Biogeochemical prospecting for gold at
99(3), 83-86. the Yellowknife City Gold Project, Northwest
DOI: https://doi.org/10.3969/j.issn.1001-4748. Territories, Canada: Part 2-Robust statistical
2007.03.026 analysis. Applied Geochemistry. 149, 105559.
[30] Wang, J., Yang, X.H., Shi, R.X., 2012. Shan DOI: https://doi.org/10.1016/j.apgeochem.2023.105559
dong sheng ren kou kong jian fen bu ge ju de [38] Ashraf, M., Loftis, J.C., Hubbard, K.G., 1997.
duo chi du fen xi (Chinese) [Spatial distribution Application of geostatistics to evaluate partial
of the population in Shandong Province at multi- weather station networks. Agricultural and For-
scales]. Progress in Geography. 31(2), 176-182. est Meteorology. 84(3), 255-271.
DOI: https://doi.org/10.11820/dlkxjz.2012.02.006 DOI: https://doi.org/10.1016/S0168-1923(96)02358-1
[31] McDonald, J.F., 1989. Econometric studies of [39] Zhang, K., Wang, Zh.R., 2023. LTPP data-based
urban population density: A survey. Journal of investigation on asphalt pavement performance
Urban Economics. 26(3), 361-385. using geospatial hot spot analysis and decision
DOI: https://doi.org/10.1016/0094-1190(89)90009-0 tree models. International Journal of Transporta-
[32] Liu, Zh.D., Luo, Sh.F., Lin, Zh.W., et al., 2021. tion Science and Technology. 12(2), 606-627.
Shan dong sheng ren kou cheng zhen hua de DOI: https://doi.org/10.1016/j.ijtst.2022.06.007
shi kong te zheng ji kong jian xiao ying yan jiu [40] Ord, J.K., Getis, A., 1992. The analysis of spa-
(Chinese) [Study on spatial-temporal character- tial association. Geogrphical Analysis. 24(3),
istics and spatial effects of population urbaniza- 189-206.
tion in Shandong Province]. Qu Journal of Qufu DOI: https://doi.org/10.1111/j.1538-4632.1992.
Normal University (Natural Science). 47(4), tb00261.x
112-119. [41] Tan, M.H, Li, X.B., Lu, C.H., et al., 2008. Ur-
[33] Li, Ch.P., 2015. Xin xi liu shi jiao de shan dong ban population densities and their policy impli-
sheng cheng shi wang luo te zheng yan jiu (Chi- cations in China. Habitat International. 32(4),
nese) [City network structure of Shandong Prov- 471-484.
ince based on information flow]. Journal of Arid DOI: https://doi.org/10.1016/j.habitatint.2008.01.003
Land Resources and Environment. 29(12), 51- [42] Li, P., Xu, Y.T., 2011. Shan dong sheng ren
56. kou kong jian fen bu te zheng fen xi (Chinese)
DOI: https://doi.org/10.13448/j.cnki.jalre.2015.399 [Characteristic study on the spatial distribution
[34] Yang, Q., 1990. Qu yu ke liu fen bu mo xing de of Shandong Population]. Ludong University
yan jiu (Chinese) [A model for interregional trip Journal (Natural Science Edition). 27(3), 274-
distribution in China]. Acta Geographica Sinica. 278.
45(3), 264-274. DOI: https://doi.org/10.3969/j.issn.1673-8020.
DOI: https://doi.org/10.11821/xb199003002 2011.03.021
[35] Wang, F.H., 2014. Quantitative methods and [43] Mao, Q.Zh., Long, Y., Wu, K., 2015. Zhong guo
socio-economic applications in GIS. CRC Press: ren kou mi du shi kong yan bian yu cheng zhen
Boca Raton. pp. 127. hua ge ju chu tan—cong 2000 nian dao 2010
[36] Ohlert, P.L., Bach, M., Breuer, L., 2023. Accu- nian (Chinese) [Spatial-temporal changes of
racy assessment of inverse distance weighting population density and exploration on urbaniza-
interpolation of groundwater nitrate concentra- tion pattern in China: 2000-2010]. City Planning
tions in Bavaria (Germany). Environmental Sci- Review. 39(2), 38-43.
ence and Pollution Research. 30(4), 9445-9455. [44] Zhang, P.F., Xu, Y.T., 2019. Ji yu GIS de shan
DOI: https://doi.org/10.1007/s11356-022-22670-0 dong sheng ren kou fen bu shi kong ge ju ji ying

79
 Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

xiang yin su fen xi (Chinese) [Spatial temporal circle]. China Journal of Commerce. 830(7), 12-
pattern and influencing factors of population dis- 14.
tribution in Shandong Province based on GIS]. DOI: https://doi.org/10.19699/j.cnki.issn2096-
Anhui Agricultural Science Bulletin. 25(9), 158- 0298.2021.07.012
162. [47] Gao, Q., 2019. De zhou cheng zhen hua fa zhan
DOI: https://doi.org/10.3969/j.issn.1007-7731. bu ping heng bu chong fen wen ti qian xi (Chi-
2019.09.063 nese) [Analysis of unbalanced and inadequate
[45] Pang, Y., Lian, H., Wang, L., 2019. De zhou shi urbanization development in Dezhou]. Modern
xian yu cheng zhen hua shui ping ce du ji qi fen Economic Information. 14, 498.
yi yan jiu (Chinese) [Study on the measurement [48] Cai, Y.J., 2021. Ji yu duo yuan shu ju de ji nan
and dissimilarity of urbanization level of De- dou shi quan tong cheng hua jin cheng ji qi ying
zhou County]. City. 227(2), 53-60. xiang yin su yan jiu (Chinese) [Research on the
[46] Zhu, Y., 2021. Ke xue gou jian jiao dong ban process and influencing factors of urban integra-
dao jing ji quan yi ti hua zhan lue ge ju (Chinese) tion of Jinan Metropolitan Area based on multi-
[Scientifically constructing an integrated strate- source data] [Master’s thesis]. Jinan: Shandong
gic pattern of the Jiaodong Peninsula economic Normal University. p. 22.

80
Journal of Geographical Research | Volume 06 | Issue 03 | July 2023

81