Skill of operational forecast of heavy rainfall events during southwest monsoon season over India

To predict extreme weather correctly has always been a challenge for operational forecaster. Recent enhancement in observational network and development of new forecasting tools & techniques has helped in increasing the understanding of weather, eventually enabling the forecaster to provide better forecast services. In India, about 80% of its annual rainfall occurs during monsoon season (June to September) in which heavy rainfall (HR) events give major contribution in causing flood situations of varying severity in different parts of the country in almost every monsoon season. Therefore, the forecast and verification of HR events during monsoon season is of great importance due to their social & economic impact. The present study focuses on verification of operational HR forecast over India during monsoon season (June to September) using data of recent 12 years (2002 - 2013). A significant improvement in skill scores in recent years is observed for 24 hours HR events forecast over India. The improvement in the accuracy of the forecast is seen from various skill scores particularly critical success index (CSI), False Alarm Rate (F) and Heidke Skill Score (HSS), which has improved by 42%, 52% and 43% respectively since 2002.