The Food and Agriculture Organizations’ (FAO) PenmanMonteith reference evapotranspiration (ET0) index is a key parameter in hydrological and meteorological studies. Temporal and spatial variations in ET0 from 1981–2017 were investigated in the topographical rapid change zone in the Hengduan Mountains, China. The results showed a change point around the year 2000 in the area-averaged annual ET0 series. ET0 decreased and increased significantly by 3.103 mm/yrmm/year (p < 0.05) from 1981–2000 and by 3.591 mm/yrmm/year (p < 0.05) from 2001–2017, respectively. The contribution analysis shows that reduction in wind speed (Ws) was the primary driving force for the decrease in ET0 during 1981–2000 in spring, autumn, and winter, and annually, while net solar radiation (Rn) was the dominant force in summer. Reduction in relative humidity (RH) was responsible for the increase in ET0 in all seasons and for the annual scale in 2001–2017. The sensitivity analysis shows that ET0 was most sensitive to Rn, followed by RH, and air temperature (Ta) was the least sensitive of the variables. The trends of ET0 were also found to increase with elevation; we denote this as the elevation-dependence of ET0 changes. The elevation-dependence was also noted for the trends of Ws, RH, and Rn, with higher elevations showing larger changes in these parameters. In addition, the sensitivities of Rn, RH, and Ta decreased with elevation, while that of Ws increased with elevation. A comprehensive investigation into the trends of these climatic variables and their sensitivities revealed complex trends of ET0 along the elevation gradient, with typical increases with elevation over the annual scale despite the large differences in seasons. A more detailed exploration of the mechanisms causing this pattern is required.
Rainfall intensity is a key factor that influences the processes of infiltration and runoff generation on the surface soil. However, the Natural Resources Conservation Service runoff curve number (NRCS-CN) method, which is widely used to simulate direct runoff, does not consider the impact of rainfall intensity on the simulation results. Hence, this study incorporates a rainfall intensity modification factor (γ) into the Ia-S relationship in the NRCS-CN method. The results show that the modified method, the NRCS-CN-γ method, improves the efficiency, reduces the impact of variable rainfall intensity on the simulation results and reduces the relative errors caused by the changes in CN to approximately one-third of the original errors in the NRCS-CN method. Consequently, the NRCS-CN-γ method contributes to a more accurate simulation and prediction of direct runoff in monsoon regions where rainfall intensity greatly varies, especially under climate change.