Assignment Cover Sheet

Please complete and insert in the front of your assignment before submitting to
ELE.

Module Code: BEE1036

Candidate Number: 152504
Your candidate number can be located via iExeter, Student Record System (SRS),

collect candidate number for 2018/9 and it’s 6 digits long

Do you have an approved Individual Learning Plan

(ILP)
Yes / No
If Yes to above, do you require and have approval
for specific learning difficult marking guidelines (as
Yes / No
outlined in the link below)?
https://www.exeter.ac.uk/media/universityofexeter/wellbei
ng/documents/Dyslexia_Marking_Guidelines.pdf

Please provide any additional information relating

to your ILP that you would like to bring to the
attention of the marker:

Mark Feedback (paper-marking only):

Feedback:

Mark:
Climate Change Empirical Project

Introduction

With the rapid industrial growth, the term “climate change” have become a hot
potato. Given that climate change is a long-term result of different variables, in the
following sections, I would show whether climate change happens right now and the
way of measuring its extent.

1. Behaviour of land-ocean temperature anomalies in the northern hemisphere

1.1 Data usage: Q1.

Instead of using absolute temperature of regions as data, temperature anomalies are
used. Temperature anomalies is a measure of relative temperature change, giving
indication on the comparative temperature change, whether it is warmer or cooler and
its amplitude. Temperature anomalies over a period of time is based on a base period
of 30-years (1951-1980) for the GISS analysis.
In compare with temperature anomalies, absolute temperature imposes a larger
variance of data for different areas within a large region, and therefore will possibly
imply higher uncertainty. However, temperature anomalies are indication of
temperature change for a much larger region of distance up to 1000km according to
Hansen and Lebedeff(1987). Therefore, temperature anomalies were chosen over
absolute temperature.
Figures:
Q2. Figure 2

Northern hemisphere temperatures (1880–2016)(in May)

1.4
Temperature anomaly (degree Celsius)

1.2
1
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
1880
1884
1888
1892
1896
1900
1904
1908
1912
1916
1920
1924
1928
1932
1936
1940
1944
1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
2004
2008
2012
2016

Year
 Q3.

Temperature anomaly(degree Celsius) Temperature anomaly(degree Celsius)

-1
0.5
1.5
-1
0.5
1.5

0
1
2
0
1
2

-0.5
-1.5
-0.5

Figure 3.2
Figure 3.1

1880 1880
1884 1884
1888 1888
1892 1892
1896 1896
1900 1900
1904 1904
1908 1908
1912 1912
1916 1916
1920 1920
1924 1924
1928 1928
1932 1932
1936 1936
1940 1940
1944 1944
1948 1948

Year
Year
1952 1952
1956 1956
1960 1960
1964 1964

season(1880-2016)
1968 1968
1972 1972
1976 1976
1980 1980
1984 1984
1988 1988

average temperature anomaly for 2nd

1992 1992
1996 1996
average temperature anomaly for 1st season(1880-2016)

2000 2000
2004 2004
2008 2008
2012 2012
2016 2016
Temperature anomaly(degree Celsius) Temperature anomaly(degree Celsius)

-1
0.5
1.5
0.2
0.4
0.6
0.8
1.2

0
1
0
1

-0.5
-0.8
-0.6
-0.4
-0.2

Figure 3.4
Figure 3.3

1880 1880
1884 1884
1888 1888
1892 1892
1896 1896
1900 1900
1904 1904
1908 1908
1912 1912
1916 1916
1920 1920
1924 1924
1928 1928
1932 1932
1936 1936
1940 1940
1944 1944
1948 1948

Year
Year

1952 1952
1956 1956
1960 1960
1964 1964

season(1880-2016)
season(1880-2016)

1968 1968
1972 1972
1976 1976
1980 1980
1984 1984
1988
average temperature anomaly for 4th
average temperature anomaly for 3rd

1988
1992 1992
1996 1996
2000 2000
2004 2004
2008 2008
2012 2012
2016 2016
Q4
Part a
Average temperature anomaly for each year :
Figure 4

Average temperature anomaly for each year(1880-2016)

1.5
Temperature anomaly(degree Celsius))

0.5

0
1951–1980 average
-0.5

-1
1880
1884
1888
1892
1896
1900
1904
1908
1912
1916
1920
1924
1928
1932
1936
1940
1944
1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
2004
2008
2012
2016
Year

Q4b.
According to figures plotted in Q2-4a, temperature and time are generally
positively correlated.

Q5.
For chart using the same month of years as time interval (ie. Figure 2), we are able
to observe the long-term changes of average temperature for the same period of time
each year, which allows us to study about the change of temperature with the change
of year as the only variable that affect the result. This ignore the effect of the difference
in seasons, and with a more detailed data in compare to the use of year as time interval.
For charts using seasons as time interval (ie. Figure 3.1-3.4), it enables us to learn
about the pattern in temperature over time for different seasons. Not only they show the
trend of change of temperature in different time period of the year, but also comparing
the effect of different seasons on the general level of the pattern of temperature.
For chart using year as time interval (ie. Figure 4), it shows the overall general
trend of the change of temperature over time. We are able to learn the change of
temperature over time from year to year, which is inclusive to factors affecting the
pattern of temperature change shown on charts with “month” and “season” as time
interval. It gives us the overall world trend, helps constructing prediction on future
pattern for year as a whole.

Q6.
a) In terms of similarities, both of them show the same pattern at the period from
around 1900 to 2000. In both chart, the temperature over time are positively
correlated, while both of the charts have shown the increase in temperature with
accelerating deviation of temperature during year 1800 to 1900. Also, the amplitude
of fluctuation between each time intervals are similar.
There’s also some difference between them. On x-axis, Figure 1.4 covered a much
wider time period of 1000 years, about 7 times Figure 4 in Q4 has covered. In Figure
1.4, different trend have been shown. During year 1000 to 1700, instead of the
increasing trend mentioned, a decreasing trend appeared, which differ from Figure
4 in Q4. Moreover, the divergence of data observed is about 2 degree Celsius while
that in Figure 1.4 is only about 1.2 degree Celsius.

b) When comparing my chart with Figure 1.4, Figure 1.4 shows a flatter trend in the
past. The deviation was most of the time below 0 with a little fluctuation which
means the mean temperature was stable and cooler than the time 1961-1990.
Meanwhile my chart shows that half of the time the data are above the mean line
even with increasing trend. Compared with the past 1000 years, this increasing
pattern is unusual.

c) Yes, the government should absolutely be concerned about climate change. Within
the recent 100 years, the average temperature of the Northern hemisphere have
shown an accelerating increase. This situation happens every seasons and individual
months, and it’s now way above the average temperature between 1951 and 1980,
which shows the severances of the rapid increase in temperature
Part 2 : Distribution of temperatures and temperature variability over time

Q1.
Range of temperature anomaly (T) Frequency : 1951-1980 Frequency : 1981-2010
-0.3 0 0
-0.25 2 0
-0.2 7 0
-0.15 3 0
-0.1 12 1
-0.05 11 3
0 7 2
0.05 13 6
0.1 14 4
0.15 10 3
0.2 7 4
0.25 4 5
0.3 0 5
0.35 0 7
0.4 0 3
0.45 0 7
0.5 0 7
0.55 0 6
0.6 0 2
0.65 0 3
0.7 0 10
0.75 0 6
0.8 0 1
0.85 0 2
0.9 0 3
0.95 0 0
1 0 0
1.05 0 0
Q2.
a)

distribution of temperatures (1951-1980)

16
14
12
10
frequency

8
6
4
2
0
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
1.05
the range of temperature anomaly (degree Celsius)

distribution of temperatures (1981-2010)

12

10

8
frequency

0
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0

1
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95

1.05

the range of temperature anomaly (degree Celsius)

b) Distribution of temperature anomalies in 1951–1980 is more concentrated than that

in 1981-2010. While the distribution in 1981-2010 spread more even than that in 1951-
1980, so as a wider range of temperature anomaly in 1981-2010.. The distribution in
1981-2010 is slightly negatively skewed. Also, the range of temperature anomaly
shifted right as a whole from 1951-1980 to 1981-2010.
Q3.
3rd decile = - 0.1
7th decile = 0.1

Q4.
Number of anomalies that are considered ‘hot’ in 1981–2010 is 306, the percentage is
85%.
Based on decile in 1951-1980, 85% of temperature observations in 1981-2010 are
considered as “hot”, which outweigh the original 30% in 1951-1980 by about 2.8 times.
More days were considered as “hot” during 1981-2010, within 30 years same as data
from 1951-1980. Therefore, it suggests that we are experiencing hotter weather more
frequently in 1981–2010.

Q5
a)
Average: DJF MAM JJA SON
1921–1950 -0.0396667 -0.053 -0.059 0.072
1951–1980 -0.003 0.00066667 0 -0.0003333
1981–2010 0.52533333 0.511 0.402 0.43033333

Variance: DJF MAM JJA SON

1921–1950 0.05498322 0.029681 0.01984233 0.027156
1951–1980 0.04908767 0.02437956 0.01420667 0.02492989
1981–2010 0.07712489 0.07398233 0.06648267 0.10796322

b) For season DJF, the variance decreased by about 0.005 in 1951-1980 comparing to
1921-1950. While period in season MAM and JJA variance also decreased by 0.05
in first two period. Season SON experienced a decrease of about 0.003. From first
two period, it shows that temperature was less variable in later period.
However, from 1951-1980 to 1981-2010, all four seasons faced significant increase
in variance. DJF increased by 0.03; MAM’s multiplied by 3 times; JJA’s multiplied
by about 4.7 times; SON’s multiplied by about 4.3 times. Temperature appeared
to.be much more variable in later period.
Q6.
From figures in Q2, we can observe that the distribution range of temperature
anomalies expanded in later period. The distribution becomes flatter. The peak value
of distribution also negatively skewed. There’s no data left below the temperature
anomaly -1. It shows that the average temperature has increased. From findings of the
New York Times, between 2005 and 2015, two-thirds of values of temperature recorded
were in the “hot” category, and nearly 15% were in the “extremely hot”. The
distribution shown in the article is also negatively skewed and flattened. It’s said that it
might be the effect that part of the countries become warmer faster than the rest. From
tables in Q5, both average temperature anomalies and variance increased significantly
in later period. While in Q4 it was suggested that we are experiencing fot weather more
frequently, which means that it’s the increase in average temperature of a portion of the
world that cause the increase in mean value. With no clue in saying that we’re
experiencing more cold days, the temperature cannot be said as more variable.
As aforementioned, we’re experiencing hot days more frequently, or even extremely
hot days mentioned in the article. Affected by extreme weather, more climate problem
arise such as more frequent heat waves. It even affects the harvest of farm, as well as
the average water level. Therefore I would definitely advise the government to spend
more money on mitigating the effects of extreme weather events.

Part 1.3 Carbon emissions and the environment

Q1. I think this data is a reliable representation of the global atmosphere. As the
observatory is located near the summit of Mauna Loa at an altitude of 3400 meters,
away from urban area, the air content will accurately by the average content among the
world. And the observatory is surrounded by mainly lava and volcanic rock. Without
the existence of vegetation or soil, they are able to measure the background content of
CO2, not the content with natural emission of soil. The above condition create a suitable
environment to measure accurate content of CO2 with less fluctuation as if it’s
conducted in urban area, which is highly affected by human activities. With stable air
condition most of the time and it’s benefit of high altitude, I think this data is a reliable
representation of the global atmosphere.
Q2.
Trend is the representation of past and current level of CO2 which is actual data. While
the interpolated data are the estimated values that lie within the measured range of data,
used for generating smoother curve and more precise analysis to put forward.
Seasonal variation in CO2 levels is caused by the fluctuation on CO2 content from
photosynthetic activity. For plants’ growth and reproduction, CO2 is the essential
element. Plants begin their photosynthesis in the spring till summer, consuming CO2
from the atmosphere, the process decrease the content of CO2 in the atmosphere,
However, in winter, plants start to slower and reduce the consumption of CO2, so as to
save energy.

Q3.

CO2 levels of the atmosphere recorded from January, 1960

450

400
350
300
CO2 levels

250

200

150
100

50
0
1960
1961
1963
1965
1966
1968
1970
1971
1973
1975
1976
1978
1980
1981
1983
1985
1986
1988
1990
1991
1993
1995
1996
1998
2000
2001
2003
2005
2006
2008
2010
2011
2013
2015
2016
Years

Trend Interpolated

The positive slope in the chart shows that CO2 level is positively related to time.
Q4.
a)

Comparison of CO2 trend and temp anomaly, 1959-2017

430
410
390
370
CO2 level

350
330
310
290
270
250
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2
Temperature anomaly (degree Celsius)

b) The correlation coefficient between CO2 level and temperature anomaly=0.93298793

As the correlation coefficient is near +1, it means that CO2 level and temperature
anomaly have a high degree, near perfect correlation, which is positive. When either
one of the variables increase, there will also be an increase of the other variable.

Q5.
a)

Comparison of CO2 trend and temp

anomaly, January, 1959-2016
430
410
390
370
CO2 level

350
330
310
290
270
250
-1 -0.5 0 0.5 1 1.5 2
Temperature anomaly (degree Celsius)

correlation coefficient = 0.7858726

 Comparison of CO2 trend and temp
anomaly, Feburary, 1959-2016

430
410
390
370
CO2 level

350
330
310
290
270
250
-1 -0.5 0 0.5 1 1.5 2 2.5
Temperature anomaly (degree Celsius)

correlation coefficient = 0.73829096

The charts and coefficients suggest that CO2 level and temperature anomalies are
positively correlated at high degree.

b) The 2 variables measured may experience their own natural change overtime, which
their final values in different period may have a larger variance. Therefore,
correlation is not representative in time series analysis since it does not account for
correlation trends over time. Also, degree of correlation does not necessarily mean
the causation of the 2 variables. As there may be third parties variables that cause
the change of the 2 measured variable, correlation may not be able to accurately
summarize the causation between them..

Q6.
a) Spurious correlation is that when two different variables are correlated at either high
or low degree, but there’s no clue in concluding their factual linkage that the change
of one variable will cause the change of another variable.
The difference between correlation and causation is that causation include refutable
fact for the linkage between them, which can be proofed by experiments. While
correlation is what concluded from statistics data only, without scientific fact about the
factual linkage in between.
b) For example, although statistics show that temperature anomaly increase when CO2
level increase and the government says it’s the results of increasing CO2 emission,
there might also be some unseen factors that cause the increase in CO2 level. For
instance, due to urbanization of developing countries, deforestation is one of the
on-going processes nowadays, in which it severely decrease the number of plants.
The decrease in number of plants will also cause the decline in the rate of
photosynthetic activity which account for the intake of CO2 . Therefore less CO2
is being converted by plants, resulting in the increase in average CO2 concentration
in atmosphere even if human activities that emit CO2 are less active. There might
not be a direct linkage between carbon emission and increase in temperature
anomaly.

c)

The above is the chart showing US crude oil imports from Norway correlates with
drivers killed in collision with railway train. In my opinion, although these’s no obvious
linkage in between, I think there may be a third variable that link them.
Assume that the change in import of crude oil from Norway follow the trend of
the change in total import of crude oil in US. The increase in crude oil imports may be
a reflection of the increasing demand of crude oil for the production of petrol and parts
used in vehicle production. In other words, there may also be an increase in demand for
cars in the US, which may also be accounted by other unseen factors like lower
examination fee for driver license. Assume the probability of a driver killed in collision
with railway train remain constant, the increase in number of cars on road will results
in higher average number of drivers killed in collision with railway train.
Moreover, the increase in crude oil import will also increase the workload of
railway train for transportation of crude oil, which then lead to a busier schedule of
train. As a result of increased frequency of train passing through, there will be higher
average number of drivers killed in collision with railway train.
d) In an article “Is Soda A Smoking Gun For Teen Violence - Or Just Statistical
Illiteracy?” (by Trevor Butterworth),Forbes 1 , it’s said that the frequency of
intaking non-diet soda is correlated with youth violence. Up till now, there’s still
not any scientific fact that proof drinking non-diet soda will cause violence.
Whether or not the ingredient of non-diet soda will ignite violence remains
unresolved. Therefore first of all there is no observable linkage in between drinking
soda and violence. While there may be other factors that leads to the result. What
ought to be consider of causing juvenile delinquency is teen’s socioeconomic state.
Over dose of soda may reflect unbalanced lifestyle or maybe presumably the
possibility of them drinking beer, in which is also soda. These subliminal factors
can account for teen’s social behavior. Without further studies, they are called to be
spurious.

1
https://www.forbes.com/sites/trevorbutterworth/2011/11/02/is-soda-a-smoking-gun-
for-teen-violence-or-just-statistical-illiteracy/#42fbab301f1b
“Is Soda A Smoking Gun For Teen Violence - Or Just Statistical Illiteracy?” (by
Trevor Butterworth),Forbes