基于CMIP6气候变化的开都河河源区径流预测OA

Using CMIP6 climate scenarios to drive the SWAT model is an effective approach for quantitatively as-sessing the impacts of climate change on runoff.Herein,the headwaters of the Kaidu River were selected as the study area,and a SWAT model was constructed and calibrated using monthly runoff data from Dashankou Station for the period 1961-2018.Six CMIP6 climate models were selected and combined using a multimodel ensemble mean(MME),and three shared socioeconomic pathway(SSP1-2.6,SSP2-4.5,and SSP5-8.5)scenarios were ap-plied to project future runoff changes in the study area from 2026 to 2100.The results indicate that the SWAT model performed well during both the calibration and validation periods,with NSE values of 0.79 and 0.74,rela-tively high KGE values,and a PBIAS of approximately-10％.The MME demonstrated better performance than the individual climate models.From 2026 to 2100,precipitation is projected to increase slightly,while air temper-ature is expected to rise continuously,with the most pronounced warming occurring under SSP5-8.5.Interannual runoff exhibits clear scenario-dependent differences,increasing gradually under SSP1-2.6,decreasing slightly un-der SSP2-4.5,and declining significantly under SSP5-8.5.The corresponding median long-term runoff change rates are 18.67％,3.04％,and-20.05％,respectively.At a seasonal scale,spring runoff increases significantly,and the annual flood peak occurs earlier.Under SSP5-8.5,the flood season is projected to shorten,while the dry season begins earlier.These significant differences in projected runoff changes among the climate scenarios indi-cate that climate warming is a key driver of future runoff evolution in the Kaidu River headwaters.This study pro-vides a scientific basis and practical guidance for adaptive water resource management in arid regions.