Revealing the Response of Evapotranspiration and Water Use Efficiency to Vegetation Changes in Different Types of Urban Green Spaces

The quantification of urban water‒carbon change is critical to our ability to assess urban water resource effectiveness and quantify urban ecological processes. In this study, we used Sentinel-2 data combined with the PT‒JPLim and analytical water use efficiency (WUE) models to simulate monthly urban evapotranspiration (ET) with a spatial resolution of 10×10 m, analyzing the WUE and ET partitioning in the main urban area of Beijing from 2016 to 2021. Verification of modeled ET based on two flux towers showed that the PT‒JPLim model performed well (Daxing: R2 = 0.84, RMSE = 16.76 mm/month; Miyun: R2 = 0.83, RMSE = 22.10 mm/month). The annual mean ET and WUE in the main urban area of Beijing were 497 mm and 0.63 gC mm−1 H2O, respectively. The carbon sequestration rate (i.e., gross primary productivity) in the main urban area of Beijing significantly increased (0.074 gC m−2 d−1 yr−1), and the proportion of water utilization (i.e., ET) decreased (−19.87 mm yr−1) during the study period, which was reflected in an increase in WUE (0.029 gC mm−1 H2O yr−1). We extracted the ET and WUE of four urban green space types (park, square, residential, and scenic green spaces) based on point of interest data. Residential green space showed the largest increases in ET (6.2%) and WUE (30%) when influenced by a 25% increase in vegetation. We projected the ET and WUE of Beijing for the period 2022‒2100 based on CMIP6. The vegetation WUE (0.0512 g C mm−1 H2O yr−1) in Beijing was projected to increase by the greatest amount under a medium forcing scenario, whereas the water consumption (0.51 mm yr−1) showed a lesser increase. This study offers new insights into the refined simulation of urban ET and WUE, enabling the quantification of water consumption and carbon sequestration processes in urban green spaces.