IMPACT OF CLIMATE CHANGE ON CROPS

Sana Bibi , Sana Sarwar

December 2024

1 Abstract

The effects of climate change on agriculture demand innovative strategies to

support agriculture industry in maintaining and improving crop production.To
address these effects and balance them, it is essential to have advance analy-
sis tools that would help determine the interdependence of climate factors and
crop production. Climate change is a challenge that can be met with the help
of Artificial Intelligence (AI) and its machine learning approaches such as Naive
Bayes, Decision Trees, and Regression Models.The purpose of this paper is at
applying these machine learning models to assess and forecast the impact of
climate change on crop production. Naive Bayes classifier, is used for classify-
ing and predicting the performance of crops under different climatic conditions.
The decision tree algorithm is easy to interpret and can handle multivariate non-
linear relationships between variables like temperature, rainfall, and humidity.
Regression models, including multiple linear regressions, help in measuring the
impact of certain climate variables on the changes in crop production over differ-
ent time periods.These models empower agriculture industry to make informed
choices and adapt effectively to the challenges brought about by climate vari-
ability.These models are assessed using historical climate and agricultural data
to determine the capability of these models in recognizing trends, assessing risks
and suggesting adaptation measures. The outcomes of this study show how the
integration of these AI models leads to better prediction and decision making
for policy makers and farmers, hence supporting climate change adaptive agri-
culture. This paper therefore demonstrates how AI can be used to tackle the
problems faced by the agricultural sector due to climate change.

1
2 Introduction

The growing impacts of climate change on agricultural systems have become

one of the major areas of concern among both researchers and policymakers.
Climate change here refers to long-term changes in temperature, pattern, and
extremity of weather events regulated by natural and anthropogenic factors,
such as the emission of greenhouse gases. It impacts crops in terms of low
yields, changing growing seasons, increased susceptibility to infestation and dis-
ease, and shifting zones of crop viability. All these factors together present a
very serious threat to food security and economic stability, especially in regions
where agriculture is the dominant activity. To address these challenges, differ-
ent machine learning techniques are in use, such as Naive Bayes, Decision Tree,
and Regression, to model and predict the impacts of climate change on crops.
Naive Bayes is a model based on Bayes’ theorem, generally used for classification
tasks, which can predict the conditions of crops with weather data. Decision
Tree: Tree-structured algorithm that simplifies complex decision-making pro-
cesses and identifies how specific climatic variables (like temperature or rainfall)
influence crop yield. Regression models, which may be linear or nonlinear, are
quantifying the relationships between climatic variables and crop productivity;
hence, the trends and future scenarios have been illustrated. Such techniques
will be of prime importance in understanding the impact of climate change on
crops due to several reasons: they enable the stakeholders to apply data-driven
adaptation strategies, such as the choice of climate-resilient crops or adjusting
planting schedules; they also emphasize the urgent need for policymakers to take
mitigation measures against the effects of climate change in securing food for
future generations; lastly, these studies foster sustainable agriculture through
predictive insights into the environmental impacts on farming. Research into the
application of artificial intelligence in climate-agriculture interactions started to
take shape at the turn of the 20th century, with major breakthroughs in the
21st century due to increased access to climate data and computation. As the
intersection of AI and agriculture continues to evolve, these studies aim to pro-
vide innovative solutions for one of the most pressing global challenges of our
time.

3 Related Work

The most pressing issue at the moment is climate change, which has a significant
impact on the Earth’s ecosystems and natural processes. The rapid pace of cli-
mate change on Earth has been a natural and unavoidable phenomenon through-
out history, but it is largely due to human activities. Almost 0.9°C has been the
global mean temperature rise since the 19th century, due in large part to green-
house gases from industrialization and pollution along with deforestation alone.
Agriculture, which is critical to the production of food and the global economy,

2
is experiencing significant effects from climate change. Global food security is
being threatened by the rapid changes in weather patterns that are linked to
the close relationship between agriculture and climate. The World Food Pro-
gramme 2018 report revealed that crops are not growing as quickly as expected
to keep up with the population growth. The Food and Agriculture Organization
warned in 2016 that if greenhouse gas emissions and climate change persist at
present, the production of essential crops could decrease significantly by 2100
.All these climate related disturbances would not only have drastic impacts on
food distribution patterns but also on its quality and access.it will also raise
humanitarian concerns as food security is deeply entwined with public health
and will create a vicious cycle of hunger, diseases and crime[1]Therefore, it is
very likely that the main drivers of climate change (increased atmospheric CO2,
increased temperature and decreased soil moisture) could significantly affect the
population dynamics of insect pests and thus the percentage of crop losses[2]In
Asia, climate variability (temperature and rainfall) and climate-driven extremes
(flood, drought, heat stress, cold waves, and storms) have several negative im-
pacts on the agriculture sector[3]Soil organic matter has been affected negatively
under current climate variability and it will continue with climate variability.
Arid regions of the globe have been affected more because of higher temper-
ature resulting in higher decomposition of organic matter into the soil [4]The
rice-wheat cropping system, a major cropping system which fills half of the
food demand in Asia, is under threat due to climate change[4] Various climate-
driven extremes, i.e., drought, heat waves, erratic and intense rainfall patterns,
storms, floods, and emerging insect pests have adversely affected the livelihood
of the farmers.[5] It is widely accepted that the changes in temperature, rain-
fall patterns, sea water level and concentration of CO2 in the atmosphere will
have the most devastating impacts on agricultural production. [6]In Asia, cli-
mate variability (temperature and rainfall) and climate-driven extremes (flood,
drought, heat stress, cold waves, and storms) have several negative impacts
on the agriculture sector[7]FAO reports that rise in world hunger level is con-
stantly increasing since 2014 and an increment in number of undernourished
people have been observed from 804 million in 2016 to 824 million in 2017[8]
Temperature is considered one of the most important factors affecting the dis-
tribution and abundance patterns of plants due to the physiological limits of
each species[9]The main cause of global warming is increased concentrations
of greenhouse gases in the atmosphere. The most prevalent atmospheric gases
are carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O), which are
caused by many anthropogenic activities including burning off the fossil fuels
and land-use change[10]

3
4 Material and Methods

The models we used in this report are Decision tree algorithm , Regression
models (Linear Regression and Multiple Regression) and Naive Bayes algo-
rithm. The dataset includes feature such as Average Temperature, Total Pre-
cipitation,CO2 Emissions ,Year, Country,Region ,Crop type , Extreme Weather
Events, Adaptation Strategies , Economic Impact

The Decision Tree algorithm: was applied to identify impurities in a

dataset by splitting data into groups.To find patterns and make predictions
effectively.

Multiple Linear Regression: was applied to describe the relationship

between target variable and independent variables like Average Temperature
,Total Precipitation , CO2 Emissions and this help to predict impact of climate
change based on agriculture.

Naive Bayes algorithm:was used for probabilistic classification, offering

insights into data patterns and yield prediction.

Python programming language: was used to implement these algo-

rithms.

Machine Learning packages numpy , matplotlib(plotting for data visualiza-

tion), Scikit-learn and SciPy was used to implement the algorithms

Mode1

A decision tree algorithm divide data into smaller groups.It starts at the top
(root node) and splits into branches until it reaches the end (leaf nodes), where
predictions or classifications are made

In decision trees, Entropy determine uncertainty(refers to the degree of

unpredictability or lack of pattern) within the data.The higher the entropy the
more uncertainty is in the data.Entropy is calculated by this formula .
Pn
H(S) = − i=1 pi log2 pi

Where:

pi is the proportion of instances in class i in the dataset D.

Information Gain (IG) measures how much uncertainty is reduce. The

highest information gain, mean that it makes the data as pure as possible.The
information gain is calculated by this formula :

4
 P |Sv |
Information Gain(S, A) = H(S) − v∈Values(A) |S| H(Sv ) Where:

• H(S) is the entropy of the set S,

• Values(A) is the set of all possible values of attribute A,
• Sv is the subset of S where the attribute A takes value v,
• |S| is the total number of instances in set S,
• |Sv | is the number of instances in subset Sv .

At each step, the decision tree chooses the feature with the highest informa-
tion gain to split the data. This process continues until the data is divided into
groups that are as pure as possible.

Mode2

Multiple Linear regression model: to make pridection independent feature

such as Temperature , Humidity, pH value , Rainfall was used.Multiple linear
Regression Equation used was: Y = β0 + β1 X1 + β2 X2 + β3 X3 + β4 X4 + ϵ

where:

• Y is dependent variable (Crop)

• X1 is Temperature
• X2 is Humidity
• X3 is pH value
• X4 is Rainfall

β0 = The intercept of the line (i.e., where the line intersects the y-axis)
β1 = Cofficient of Temperature
β2 = Cofficient of Humidity
β3 = Cofficient of pH value
β4 = Cofficient of Rainfall

5 Result and Discussion

The Decision tree and Regression model was implemented on the CSV File.

5
 Figure 1: Decision Tree

Figure 2: Caption

There are seven columns . Nitrogen, Phosphorus,Potassium,Temperature,

Humidity, pH Value, Rainfall

Target variable crop.

6 Experiment and Innovation Interpretation

6.1 Decision Tree Result The decision tree algorithm achieved an accuracy
of 98.33 percentage on the dataset. it shows that the model correctly classified
98.33 percentage of data .

Accuracy: 0.9833333333333333

6.2 Multiple Linear Regression result

The decision tree model’s performance was evaluated using two key metrics:
Mean Squared Error (MSE) and R2 score. The model achieved an MSE of 28.59

6
which measures the average squared difference between the predicted and actual
values. A lower MSE generally indicates better predictions and in this case the
error suggests room for improvement.The R2 score which was 0.305 indicates
that the model explains about 30.5 percent of the variance in the target variable.

Mean Squared Error: 28.59485653390435

R2 Score : 0.3047976967664494

When:

Nitrogen=108 , Phosphorus=94, Potassium=47 , Temperature=27.359116,

Humidity=84.54625, pHV alue =6.3874314Rainf all =90.812505

Output:

Predicted Crop: Coffee

References
[1] Naveen Kumar Arora. Impact of climate change on agriculture produc-
tion and its sustainable solutions. Environmental sustainability, 2(2):95–96,
2019.
[2] Babasaheb B Fand, Ankush L Kamble, and Mahesh Kumar. Will climate
change pose serious threat to crop pest management: A critical review.
International journal of scientific and Research publications, 2(11):1–14,
2012.
[3] Food and Agriculture Organization. The impact of disasters on agriculture:
Addressing the information gap. 2017.
[4] Abdul Ghaffar, Muhammad Habib Ur Rahman, Saeed Ahmed, Ghulam
Haider, Ishfaq Ahmad, Mehmood Alam Khan, Muhammad Afzaal, Shakeel
Ahmed, Shah Fahad, Jamshad Hussain, et al. Adaptations in cropping
system and pattern for sustainable crops production under climate change
scenarios. In Improvement of plant production in the era of climate change,
pages 1–34. CRC Press, 2022.
[5] Muhammad Habib-ur Rahman, Ashfaq Ahmad, Ahsan Raza, Muham-
mad Usama Hasnain, Hesham F Alharby, Yahya M Alzahrani, Atif A
Bamagoos, Khalid Rehman Hakeem, Saeed Ahmad, Wajid Nasim, et al.
Impact of climate change on agricultural production; issues, challenges,
and opportunities in asia. Frontiers in Plant Science, 13:925548, 2022.
[6] Benjamin Kipkemboi Kogo, Lalit Kumar, and Richard Koech. Climate
change and variability in kenya: a review of impacts on agriculture and

7
 food security. Environment, development and sustainability, 23(1):23–43,
2021.
[7] Shelley McGuire. Fao, ifad, and wfp. the state of food insecurity in the
world 2015: meeting the 2015 international hunger targets: taking stock
of uneven progress. rome: Fao, 2015. Advances in Nutrition, 6(5):623–624,
2015.
[8] Ali Raza, Ali Razzaq, Sundas Saher Mehmood, Xiling Zou, Xuekun Zhang,
Yan Lv, and Jinsong Xu. Impact of climate change on crops adaptation
and strategies to tackle its outcome: A review. Plants, 8(2):34, 2019.

[9] Sandra Skendžić, Monika Zovko, Ivana Pajač Živković, Vinko Lešić, and
Darija Lemić. The impact of climate change on agricultural insect pests.
Insects, 12(5):440, 2021.
[10] Kelvin O Yoro and Michael O Daramola. Co2 emission sources, greenhouse
gases, and the global warming effect. In Advances in carbon capture, pages
3–28. Elsevier, 2020.