An Enhanced Bottom-Up Approach to Assess the Catchments’ Vulnerability to Climate Change

Managing water resources is becoming difficult due to projection uncertainty in the future climate. Traditional top-down modeling paradigms do not provide sufficient information to water resource managers for proper decision-making in the present uncertainties. It is due to huge uncertainty in the future projection of GCM and uncertainties in GCM downscaling techniques parameterization of hydrological models. We built upon a recently developed bottom-up approach to estimate the catchments' vulnerability to climate change. The framework identifies the vulnerable combination of climate (temperature and precipitation) that causes adverse conditions for an indicator of vulnerability, such as depletion of water or less water availability. In this work, we specify the vulnerability of the catchment by the reduction in the mean annual runoff by 50% of its historical flow. We simulate synthetic climate scenarios using historical data of the past few decades with three techniques, Synthpop, KNN-CAD, and Maximum Entropy Bootstrap Weather Generator (MEBWG), which can best simulate the frequency and extrema event of the future climate. Synthetic climate data were used to compute the indicators that we divided into the vulnerability classes. A data mining algorithm [classification and regression tree (CART)] is used to associate the vulnerability classes with the synthetic climate combinations. We use a total of 77 US catchment data in the study and compute the critical threshold for each catchment for climatic variables (precipitation, temperature, and land use). The critical threshold provides an estimation of catchments' vulnerability to climate change and can be used for critical decision-making under uncertainties for water policies.