Causal Teleconnection between Pacific Decadal Oscillation and Droughts over India

Droughts are recurring natural phenomena having devastating impacts on the ecosystem and agro-economics in almost every part of the world. The hazard caused by drought costs more than any other natural calamity. Better prediction of droughts may result in better preparedness which, in turn, helps in reducing the hazard caused by droughts. In many parts of the world, drought events have been found to possess a link with the Pacific Decadal Oscillation (PDO). A better understanding of this relationship may help us in improving the drought prediction models. However, for India, the nature of this relation largely remained unexplored. Therefore, to quantify the causal relationship between PDO and droughts in India, a Granger Causality test-based methodology has been explored. Results of the linear Granger Causality test have been compared against a recently developed neural-network-based nonlinear Granger causality-based test for Standard Precipitation Evapotranspiration Index (SPEI) at 3, 6, 9, and 12-month scales.  The results of this study suggest the significant causal teleconnections between PDO and droughts in most parts of India at all analyzed scales. Further, it was found that the nonlinear model is able to capture significant causality for more parts of the country than the traditional linear model.