Abstract
The newly developed high resolution (12‐km) reanalysis products under the Indian Monsoon Data Assimilation and Analysis (IMDAA) project of the National Centre for Medium Range Weather Forecasting (NCMRWF), India, could improve drought monitoring capabilities at a local scale. For verifying this hypothesis, the performance of IMDAA‐based drought assessment needs to be explored in terms of capturing variability of local hydrometeorological drought characteristics. In this context, the present study examined the: (i) characteristics of meteorological droughts evaluated using IMDAA rainfall, and (ii) those of hydrological droughts evaluated using flows simulated by a hydrological model Soil and Water Assessment Tool (SWAT) driven by IMDAA inputs. Drought monitoring at subbasin‐scale is performed over the Govindpur watershed lying in the Subarnarekha‐Brahmani‐Baitarani River Basin in eastern India, using popular drought indices: standardised precipitation index (SPI) and standardised streamflow index (SSI). Our results suggest that the streamflow response of the watershed predicted by the hydrological model is slightly better when high resolution inputs are utilised. Hydrometeorological drought characteristics (duration, severity, number of events) computed using IMDAA products could capture distinct patterns in western and southern subbasins of the study area, while identifying drought hotspots. Comparison of drought characteristics of two historical time slices (1981–2000; 2001–2020) considered in this study suggests that severity and spatial variability of drought events have reduced in recent years. In the historical analysis, IMDAA‐based drought early warning for lead times up to 2 months has also shown promising results for detecting hydrological drought events. In our study area, drought propagation from rainfall deficit to hydrological droughts at a subbasin scale would occur faster (1–3 months) in the summer season, while it took longer (4–12 months) in the other seasons.