Multi-method attribution of the extreme precipitation and flood of 2018 in Kerala, India.

The role of global warming and climate change in altering the likelihood of extreme weather events is becoming increasingly evident. Event attribution refers to the collection of frameworks that use observed data and climate model simulations for quantifying the contribution of human-induced (anthropogenic) climate change in changing the event probability. In this study, we present a multi-method event attribution analysis of the catastrophic extreme precipitation and flooding event in Kerala, India in August 2018, that resulted in widespread destruction and loss of lives. Two methods- (i) based on factual (Historical) and counterfactual (HistoricalNat) runs from 5 CMIP6 climate models, and (ii) based on observed data, scaled to 2018 (factual) and 1901(pseudo-counterfactual) climates, are considered for quantifying the fraction of attributable risk (FAR) of the 2018 event. Using an objective approach, we first define the 2018 event as the 4-day cumulative rainfall over the Periyar river basin (PRB), during 15- 18 August, 2018. This event has a return period of 373 years (90% CI: 72-1200 years). The 1-day maximum streamflow at one of the outlets of the PRB, where maximum impact during the event was reported, is used for attributing the associated flood event. Simulated using VIC hydrological model, the streamflow event is found to have a return period of 34 years (90% CI: 12-286 years). The FAR from the climate model ensembles is -0.18 and -0.14, for the precipitation and streamflow events, respectively. The scaled observations also give negative FARs: -0.97 for precipitation and -0.93 for streamflow. These values imply that the 2018 event is exceptionally less likely due to climate change. In other words, our results underline the definitive absence of anthropogenic role in the 2018 event.