Doshi Yash Lokeshkumar

GE4050

PGE21410

Indian agriculture depends on the monsoon, providing most of the country's annual rainfall. About 60% of the net sown area in India is rain-fed, which means it relies on monsoon rainfall for crop production. The monsoon season typically lasts from June to September and provides about 70% of the total annual rainfall in India. There has been a much erratic rainfall pattern in recent years. Erratic rainfall is a phenomenon that refers to unpredictable variations in the timing, amount, and intensity of rainfall. It is a significant challenge for agriculture, as crops and livestock depend on consistent and reliable water availability for growth and productivity. This research uses an Advance onset of monsoon dates from 2006-2020 with Actual and normal dates of these years for villages over India. The training dataset is produced using the same plot-level data of major crops in India to estimate the impact of early-season rainfall deficits and a Convolution neural network model that is good at identifying relationships in temporal datasets to predict crop losses using rainfall data from the first 30 days of the crop season. The trend in rainfall shows that most of the years indicate negative trends despite the year 2019. Also, it has been observed that rainfall, temperature, soil moisture and soil type have significant effects on crop yield. 


Report Content

Need for Study

Introduction

Study Area

Datasets

Methodology

Statewise Monsoon periods

TImeseries Rainfall Analysis

Computin State Wise Yield

Implementation of Convolution Neural Network

Conculsion & Limitation