Comprehending the impact of hydro-meteorological droughts on ecosystem vulnerability and resilience across the Indus...

This study develops a catchment-based integrated drought index (CIDI) across the Indus River Basin of Pakistan. CIDI is developed by systematically integrating the Standardized Precipitation Evapotranspiration Index (SPEI) and the Standardized Water Availability Index (SWAI). Initially, Bayesian Vector Auto-Regressive is used to select sensitive agricultural drought indices against SPEI and SWAI, where Bayesian Principal Component Analysis (BPCA) is employed to develop the Integrated Agricultural Drought Indices based on SPEI (SPEI-IADI) and SWAI (SWAI-IADI). Finally, Bayesian Model Averaging is applied to develop CIDI. The CIDI, SPEI-IADI and SWAI-IADI are comprehensively evaluated across 27 catchments within the Indus River Basin. The ecosystem vulnerability and resilience to drought is assessed using the Remote Sensing-based Ecological Index during 1990-2023 at 0.25° × 0.25°. The results show insignificant causal relationship of SPEI with Standardized Soil Moisture Index (SSMI) and Vegetation Condition Index (VCI) in the Upper Indus Basin (UIB); Temperature Condition Index (TCI), SSMI, and arithmetic mean-based Vegetation Health Index (VHIa) in Middle Indus Basin (MIB), and SSMI in the Lower Indus Basin (LIB). Similarly, SWAI depicted insignificant causal relationship with SSMI, TCI, and VHIa across UIB. The relationship between SWAI and SSMI across middle and lower Indus Basins is relatively dominant as compared to that with SPEI. BPCA highlights the significant role of temperature in the onset, spatiotemporal propagation, and transition of droughts. However, some pixels in MIB depicted significant variations in BPCA weights dominated both by temperature- and vegetation-derived indices. Bayesian Model Averaging assigned minimal weights to SWAI-IADI, likely due to its shorter duration and propagation time. CIDI proved to be a more robust index, outperforming both SPEI-IADI and SWAI-IADI. Further, ecosystem is found extremely vulnerable in LIB and MIB both in space and time. In contrast, high resilience is observed in UIB, attributed to its favorable cryosphere and climate. Overall, the findings suggest the importance of integrating ecosystem vulnerability and considering contributions from meteorological and hydrological droughts when developing integrated indices.