Meteorological Drought Assessment in India using Machine Learning Techniques

Global climate change causes a decrease in precipitation in India, as in many other places around the globe. As a result, droughts have occurred over a large area and in a more drastic way than in the past. The present study focuses upon assessment of meteorological drought conditions with Machine learning algorithms, such as Support Vector Machine (SVM) and Decision Tree (DT) based upon the measurement of precipitation over 36 meteorological -subdivisions in India. Meteorological drought occurs when precipitation is less than average over a prolonged time scale. It usually precedes various kinds of drought and adversely affects the environmental, economic and social conditions. Among various methods for assessment of meteorological drought by using drought indices, Machine learning algorithms can be of vital use in evaluating systems. Operating on multi-sensory data sets like SVM and DT, we will have far-reaching impacts to overcome socio- economic vulnerability in India. However, this preliminary study is focused on precipitation impacting meteorological drought. Building on the Machine learning algorithms, it is now possible to timely conduct real-time evaluation by SVM and DT applications. This approach will help us to have a refined view of drought transformations. Timely comprehensive analysis, severity and trends of meteorological drought category. Other drought interactions that are still relatively unknown, particularly in India, are to be more focused.