In this project, we predict the probability of occurrence of risk in the Indian territory, based on the historical data of precipitation and temperature between the years 1951 - 2015. This is obtained by using the Multiple Logistic Regression (MLR) model, followed by calculation of joint probability using the Copula methodolgy.
A number of experiments have been performed for measuring the probability of risk in different regions, based on multiple thresholds of drought index values and temperature. Drought indices are classified as moderate, severe and extreme drought based on their SPEI values, and temperature have been classified with different percentile values.
The risk is estimated for two-time scales: monsoon season and Water year, followed by a trend analysis depicting it's change over the years.