i

Comparative Analysis of Drought Impact on Soil

Nutrients using Deep Learning Models
A MINOR PROJECT REPORT

Submitted by

THALISETTI VENKATA PRADEEP [Reg No:RA2011028010158]

BANDARU ANAND KUMAR [Reg No: RA2011028010156]

Under the Guidance of

Dr. TYJ NAGA MALLESWARI

Associate Professor, Department of Networking and Communications

In partial fulfillment of the Requirements for the Degree of

BACHELOR OF TECHNOLOGY
in
COMPUTER SCIENCE ENGINEERING
With a specialization in CLOUD COMPUTING

DEPARTMENT OF NETWORKING AND COMMUNICATIONS

SCHOOL OF COMPUTING
COLLEGE OF ENGINEERING AND TECHNOLOGY
SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
KATTANKULATHUR -603203
NOVEMBER 2023
 ii

SRM INSTITUTE OF SCIENCE AND TECHNOLOGY

KATTANKULATHUR – 603 203

BONAFIDE CERTIFICATE

Certified that this project report titled “Comparative Analysis of drought impact on soil nutrients using
deep learning models” is the bonafide work of Mr. Thalisetti Venkata Pradeep [RA2011028010158]

and Bandaru Anand Kumar [RA2011028010156] who carried out the project work under my
supervision. Certified further, that to the best of my knowledge the work reported herein does not
form part of any other thesis or dissertation on the basis of which a degree or award was conferred
on an earlier occasion for this or any other candidate.

Dr. TY J NAGA MALLESWARI Dr. K. ANNAPURANI

Supervisor and Associate Professor Professor and Head of the Department
Dept. of Networking and Communications Dept. of Networking and Communications
SRM Institute of Science and Technology SRM Institute of Science and Technology

Signature of Examiner I Signature of Examiner II

 iii

Department of Networking and Communications

SRM Institute of Science & Technology
Own Work Declaration Form
Degree/ Course : Bachelor of Technology, Computer Science Engineering

Student Name : Thalisetti Venkata Pradeep, Bandaru Anand Kumar

Registration Number : RA2011028010158, RA2011028010156

Title of Work : Comparative Analysis of Drought Impact on Soil Nutrients

using Deep Learning Models
We hereby certify that this assessment compiles with the University’s Rules and Regulations
relating to Academic misconduct and plagiarism, as listed in the University Website,
Regulations, and the Education Committee guidelines.
We confirm that all the work contained in this assessment is my / our own except where
indicated, and that we have met the following conditions:
• Clearly references / listed all sources as appropriate
• Referenced and put in inverted commas all quoted text (from books, web, etc)
• Given the sources of all pictures, data etc. that are not my own
• Not made any use of the report(s) or essay(s) of any other student(s) either past or present
• Acknowledged in appropriate places any help that I have received from others
(e.g.fellow students, technicians, statisticians, external sources)
• Compiled with any other plagiarism criteria specified in the Course handbook /
University website
I understand that any false claim for this work will be penalized in accordance with
the University policies and regulations.

DECLARATION:

We are aware of and understand the University’s policy on Academic misconduct and plagiarism
and we certify that this assessment is our own work, except were indicated by referring, and that
we have followed the good academic practices noted above.

Student 1 Signature:

Student 2 Signature:

Date:
If you are working in a group, please write your registration numbers and sign with the date for every student in your
group.
 iv

ACKNOWLEDGEMENT

We express our humble gratitude to Dr C. Muthamizhchelvan, Vice-Chancellor, SRM Institute of

Science and Technology, for the facilities extended for the project work and his continued support.

We extend our sincere thanks to Dean-CET, SRM Institute of Science and Technology, Dr T.V.Gopal,
for his invaluable support.

We wish to thank Dr Revathi Venkataraman, Professor & Chairperson, School of Computing, SRM
Institute of Science and Technology, for her support throughout the project work.

We are incredibly grateful to our Head of the Department, Dr. Annapurani K, Professor, Department
of Networking and Communications, SRM Institute of Science and Technology, for her suggestions
and encouragement at all the stages of the project work.

We want to convey our thanks to our program coordinator Dr. G. Suseela, Associate Professor, Panel
Head, Department of Networking and Communications, SRM Institute of Science and Technology, for
her inputs during the project reviews and support.

We register our immeasurable thanks to our Faculty Advisor, Dr. N. Senthamarai, Assistant Professor,
Department of Networking and Communications, SRM Institute of Science and Technology, for leading
and helping us to complete our course.

Our inexpressible respect and thanks to our guide, Dr. TYJ Naga Malleswari, Associate Professor,
Department of Networking and Communications, SRM Institute of Science and Technology, for
providing us with an opportunity to pursue our project under her mentorship. She provided me/us with
the freedom and support to explore the research topics of our interest. Her passion for solving problems
and making a difference in the world has always been inspiring.

We sincerely thank the Networking and Communications Department, staff and students, SRM Institute
of Science and Technology, for their help during our project. Finally, we would like to thank parents,
family members, and friends for their unconditional love, constant support, and encouragement.

THALISETTI VENKATA PRADEEP [RA2011028010158]

BANDARU ANAND KUMAR [RA2011028010156]
 v

ABSTRACT

An analysis of the comparative effects of drought on soil nutrients is presented in this abstract.
Tabular fit provides a thorough investigation of the complex interactions between soil nutrient
dynamics and drought using Deep Learning Models such as LSTM, CNN, and ARIMA.
Recurrent natural disasters like drought, which are brought on by less rainfall, provide serious
obstacles to economic growth, environmental stability, and agriculture. The four different deep
learning models in this project—Long Short-Term Memory (LSTM), Convolutional Neural
Networks (CNN), ARIMA, and Tabular fit—are systematically compared in order to overcome
these issues.
We want to know how well these models work and how accurate they are in predicting how
drought would affect soil nutrient levels in different parts of the world. We face and effectively
navigate a number of obstacles during the project, such as data harmonization, computational
resource distribution, parameter sensitivity, overfitting, and data constraints. These difficulties
deepen our understanding of the field and strengthen the validity of our conclusions. The
project's results provide insightful information about the benefits and drawbacks of deep
learning models for analyzing soil nutrients during droughts, thereby equipping environmental
scientists, disaster management, and agriculture stakeholders with useful suggestions. This
research also emphasizes the value of data-driven strategies for comprehending and resolving
complicated environmental problems, encouraging sustainable practices, and building a more
resilient future.
As we get to the end of this project, we understand that finding new information and creative
solutions to environmental problems is a continuous process. The conclusions and methods
discussed here provide as a starting point for additional study and advancement in the
discipline, advancing our understanding of and ability to respond to the urgent problems of soil
nutrient management and drought.
 vi

TABLE OF CONTENTS

Chapter TITLE Page

No.
Abstract v
List of Tables ix
List of Figures x
List of Abbreviations xi

1. INTRODUCTION 12

1.1 Types of Droughts 13

1.2 Standard Precipitation Index (SPI) 16
1.3 Human Activities 17
1.4 Effects on Soil Nutrients 19
1.5 Deep Learning Models for Drought Prediction 21

2. LITERATURE REVIEW 24

2.1 Statistical Method 24

2.2 Neural network techniques 24
2.3 Other approaches
2.3.1 Atmospheric electricity 26
2.3.2 Machine learning methods 26

3. PROPOSED METHODOLOGY 28

3.1 Methodology 28
3.1.1 Data Collection and Preprocessing 28
3.1.2 Model Selection and Development 29
3.1.3 Training and Validation 29
3.1.4 Data Fusion and Interpretation 29
 vii

3.1.5 Performance Evaluation 29

3.1.6 Recommendations and Reporting 30
3.2 Standard Precipitation Index (SPI) 30
3.3 Long Short Time Memory (LSTM) 34
3.4 Convolutional Neural Network (CNN) 38
3.5 Autoregressive Integrated Moving Average (ARIMA) 40
3.6 Tabular fit or PyTorch Tabular 41
3.7 Hosting in AWS 43

4. TECHNICAL REUIREMENTS 51

4.1 Software Requirements 51

4.2 Hardware Requirements 52

5. CODING AND TESTING 54

5.1 SPI (Code) 54

5.2 LSTM (Code) 56
5.3 CNN (Code) 60
5.4 ARIMA (Code) 64
5.5 PyTorch Tabular (Code) 67

6. RESULTS AND ANALYSIS 73

6.1 SPI results 73

6.2 LSTM results 74
6.3 CNN results 75
6.4 ARIMA results 76
6.5 PYTORCH TABULAR results 77

7. CONCLUSION AND FUTURE SCOPE 80

REFERENCES 83
 viii

APPENDIX 86
PAPER PUBLICATION STATUS 89
PLAGIARISM REPORT 90
 ix

List of Tables

1.1 Nutrient Values 20

3.1 Sample Dataset Values 28

6.1 LSTM model actual vs predicted values 74

6.2 CNN model actual vs predicted values 75

6.3 ARIMA model actual vs predicted values 76

6.4 PyTorch Tabular model actual vs predicted values 77

6.5 Comparison between models 78

 x

List of Figures
Fig.No TITLE PAGE NO

1.1 Overview of Drought Classifications 15

1.2 Classification of Drought based on the SPI values 17

1.3 Effects of Human Activities on Drought occurrence 18

1.4.1 pH values 20

1.4.2 Total Soluble Salts 21

1.4.3 Deep Learning Models for Drought Prediction 23

3.3.1 Closed Loop Forecasting of LSTM (example) 35

3.3.2 Block Diagram of LSTM 37

3.3.3 Block Diagram of CNN 40

3.4 Block Diagram of ARIMA 41

3.5 Block Diagram of PyTorch Tabular 43

6.1 SPI Calculation for different timelines(3, 6, 9, 12, 24 months) 74

6.2 Plots of LSTM actual vs predicted values 75

6.3 Plots of CNN actual vs predicted values 76

6.4 Plots of ARIMA actual vs predicted values 77

6.5 Plots of PyTorch Tabular actual vs predicted values 78

 xi

LIST OF ABBREVATIONS

ARIMA Autoregressive integrated moving average

AWS Amazon Web Services

CNN Convolutional Neural Networks

DSI Drought Severity Index

EC2 Elastic Compute Cloud

LSTM Long Short-Term Memory

MSE Mean Square Error

NDMC National Drought Mitigation Center

PET Potential Evapotranspiration

pH Potential of Hydrogen

RNN Recurrent Neural Network

RMSE Root Mean Square Error

R-SQUARED Coefficient of Determination

SOM Soil Organic matter Decomposition

SPI Standard Precipitation Index

SPEI Standardized Precipitation Evapotranspiration Index

 12

CHAPTER 1

INTRODUCTION

A large amount of the 22% of global natural disaster-related economic loss is attributable to
drought, which is frequently cited as one of the most economically destructive natural disasters on
Earth. In order to evaluate ecosystem health and lessen the negative consequences of extreme
weather events on human society, drought monitoring and forecasting are crucial. A
meteorological drought, which is defined by a notable decrease in precipitation, lowers the amount
of water available on land, impacting vital elements including rivers, lakes, snowpack, and ice
sheets. The reduction in water availability has a significant effect on the soil, interfering with the
chemical, physical, and biological processes that are necessary for the health of the soil and plants.
The impact of meteorological drought on plant nutrient availability is one of its main aftereffects.
Water is the principal carrier via which vital nutrients are transported into plants; thus, a shortage
of it can impede crops' ability to absorb nutrients, impacting a variety of biological and chemical
processes during the growth season. Furthermore, the lack of moisture can cause the soil to warm
up, which might change microbial activity and nutrient processing—two processes that are
essential for plant biomass and grain production. The Soil Organic Matter Decomposition (SOM)
process may be impacted by this temperature shift in the soil, which could result in a greater release
of carbon dioxide into the sky.
Drought conditions can also cause nutrients like nitrogen to be released into the soil as nitrate.
Ironically, plants might not profit from this increased nutrient availability since the dryness inhibits
plants from absorbing the nutrients, which causes changes in biological and chemical processes
during the growing season. These modifications have wide-ranging effects on crop yield, quantity,
and a host of other interactions, such as changes in soil chemical concentrations and enzyme
activity.
Drought propagation is significantly influenced by human activity. Drought dynamics can be
greatly affected by various activities, including but not limited to dam construction, irrigation,
urbanisation, and changes in land cover and use. For example, by controlling runoff, reservoir
development plays a major role in altering the way that drought spreads. The dynamics of drought