Analysis of persistence in the flood timing and the role of catchment wetness on flood generation in a large river basin in India

Mahanadi River Basin (MRB) is one of the largest tropical pluvial river basin systems in India contributing the major source of freshwater to more than 71 million people in east-central India. Being located in the monsoon "core" region (18-28 degrees N latitude and 73-82 degrees E longitude) and its proximity to Bay of Bengal, Mahanadi River Basin (MRB) system is highly vulnerable to tropical depression-induced severe storms and extreme precipitation-induced fluvial floods during southwest monsoon as reflected in several successive and catastrophic flood episodes in recent years (2001, 2003, 2006, 2008, 2011, 2013, 2014, 2016, 2019). While previous studies so far focused on analyzing either flood trends or frequency and show the role of precipitation in flood generating mechanism over MRB using both instrumental records and climate model simulations, this study for the first time examines space-time coherence in floods and the role catchment wetness in flood response (i.e., magnitude and the timing of floods) over the basin. We examine the incidence of flooding in three different time windows: 1970-2016 (whole 47 years), 1970-2006, and 2007-2016 (post-2007s) using monsoonal maxima peak discharge (MMPD) and peak over threshold (POT) series at 24 stream gauges spatially distributed over the basin. Our analysis reveals the mean dates of floods for most of the gauges are temporally clustered during the month of August irrespective of the type of flood series and the choice of time frames. Further, we observe sensitiveness of runoff responses (flood magnitude, FM and the flood timing, FT) to lagged d-day mean catchment wetness (CW), suggesting a physical association between them. We also note FT is more strongly correlated (as manifested by statistically significant correlations) to CW rather than FM. Overall, we observe, the correlation of CW versus FT is negative, where the flood timing is relatively irregular. The outcomes of the study help to improve the predictability of floods, which can, in turn, enhance existing flood warning techniques.