Critical influence of vegetation response to rising CO2 on runoff changes.

Satellite observations show widespread greening over the global land, which potentially contributes to runoff (R) changes. However, there are discrepancies in the impacts of vegetation greening on R under elevated atmospheric CO2 concentration (eCO2). Here, we proposed an improved Budyko framework with full consideration of the vegetation structural (STR) effect and vegetation physiological (PHY) effect and evaluated runoff changes (ΔR) due to eCO2-induced vegetation variations. We found a better performance of the improved Budyko framework in simulating runoff changes from global climate models (the Nash-Sutcliffe efficiency coefficient (NSE) is 0.82). However, ΔR would be overestimated (underestimated) by 188 % (165 %) when considering the PHY (STR) effect only. Attribution analyses indicated that PHY and STR effects contribute to a ΔR of 12.8 % and - 62 %, respectively, suggesting that PHY and STR effects are indispensable variables in the projection of ΔR. The contribution of the STR effect to future ΔR is 4.8 times larger than the PHY effect, leading to a negative effect of vegetation changes on ΔR in response to eCO2. While the magnitude of PHY appears less than that of STR, the influence of PHY on ΔR follows a faster-increasing tendency in future R projections when compared to STR. Our findings emphasize the critical influence of vegetation response to eCO2 in future R projection.