Evaluating climate change impacts on the hydrology of watershed in northwestern China using a stepwise‐clustered downscaling approach

ABSTRACT In this study, a stepwise‐clustered downscaling model ( SCDM ) is advanced for transferring atmospheric simulation outputs to acquire high‐resolution climate projections at a large‐scale watershed system. SCDM can operate different temporal resolutions of atmospheric variables with continuous and discrete complexities. SCDM coupling with hydrological model is used for evaluating climate change impacts on hydrology of the Kaidu watershed in northwestern China. The daily and monthly series of large‐scale atmospheric simulation outputs for the Kaidu watershed are extracted from the ensemble of GCMs during past, recent and future periods. Results reveal that (1) SCDM is capable of downscaling climate projections for different stations, and can help understand the spatial heterogeneity of climate change, (2) the performance of SCDM is more acceptable for temperature than precipitation, (3) increase trends of T min and T max (minimum and maximum temperatures) from recent to future are projected. Besides, results from multiple downscaled climate change projections are used for driving a daily climate‐streamflow hydrological model. Results disclose that the streamflow would increase because temperature change will cause more glacier melt in future.