Compounding Risk of Heat Stress-Rain Induced Floods in Urban India

The Indian subcontinent is one of the hotspots of deadly heat stress. Several attribution studies have shown increasing trends in heatwaves and its linkage with dry spells over South Asia. However, very few studies have investigated concurrent or successive (lagged d-day) occurrence of humid heat stress (high temperature compounded by humidity) and precipitation extremes within a short time window. Using gauge-based observation records of the last five decades, we have analyzed the concurrence of extreme wet bulb temperature, Tw and peak rain events in 9 urban locations of India, distributed over climatologically heterogeneous regions. We find a larger fraction of the population is exposed to a significant increase (more than 1% and up to 2.5%/decade) in mean and extreme Tw (around 1%/decade) in several sites than solely accounting dry-bulb temperature trends. This prompted us to analyze the compound hazard associated with storm events preceded by extreme Tw (assessed through 95 98.5th percentile exceedances and annual maxima series) up to a week of occurrence of the event. Considering synchronicity between two drivers (extreme Tw and peak rain), we demonstrate cities located across the western half of the country showed positive dependence, whereas those located over the eastern half show negative dependence. While negative correlation suggests the concurrence of dry and hot episodes, the positive correlation suggests robust amplification in precipitation extremes. This is confirmed by the large upper tail distributions of peak rain events during the core monsoon season (June September) to locations showing positive dependence. Based on extreme Tw-precipitation sensitivity, we propose compound heat stress rain-induced flood hazard model for densely populated areas. Understanding drivers of peak runoff responses would benefit risk management, insurance, and flash flood forecast, devising flood resilience under climate change.