Study to Forecast Drought Using SARIMA Model

Abstract

Drought is an unpredictable disaster with a cumulative nature. Its occurrence, persistence, end, and severity cannot be characterized or formulated through any single parameter. Drought is a complex phenomenon, and nobody can predict its onset, persistence, severity, or end accurately.

In our research area of Madhya Pradesh, 90% of the rainfall occurs due to the southwest Monsoon, but irregular, low rain, and early withdrawal of monsoon can cause drought-like conditions in the area. This drought problem is more severe in arid and semi-arid zones like the Bundelkhand area.[5] The frequent occurrence of dry conditions in a particular region has an adverse impact on many sectors, especially agriculture.

Several models and indices have been developed to understand drought characteristics, but no single model is entirely effective at predicting them accurately. SARIMA (Seasonal ARIMA) is one of the most powerful and popular statistical models used to predict monthly time series precipitation data.