资源环境科技发展态势分析平台

We applied statistical and numerical modeling approach to evaluate the sensitivity of runoff (ROF) to climate variables using Global Land Data Assimilation System (GLDAS) data and regional climate model (RegCM4). It was observed that ROF is more sensitive to precipitation (P) compared to other analyzed hydroclimatic variables (potential evapotranspiration (PET), 2m air temperature (T2m), solar radiation (Rn), specific humidity (SSH), and wind speed (U), especially over India, Indochina, and south-north-northeast China semihumid-humid climate transition zones based on the higher correlation coefficient (>0.7) and elasticity (>2). The abnormal positive T2m-ROF observed over Tibetan Plateau region (TP) may be due to its high topography and cold weather regime, while positive PET-ROF over India and north China-southeast Mongolia regions can be attributed to the stronger influence of local land-atmosphere interactions. Soil moisture (SM) reflects high correlation with runoff, especially over the climate transition zones (i.e., India and Indochina-southeast China). The initial wet (dry) soil moisture (SM) anomalies lead to an increase (decrease) of ROF in each season with the hot spots mainly located in middle to high latitudes (spring), TP and northeast (summer and autumn), and Indochina (autumn) regions. Such influence can persist almost 4 months in spring while only about 1 month in autumn during dry and wet conditions. The wet condition has stronger influence at beginning but dissipates quickly, while the dry condition can last longer within the same season. The impact of initial soil temperature anomalies on ROF is weaker than SM, with the only obvious ROF changes located over south China (spring and summer) and north India (autumn).