Effect of leaf area index change on evapotranspiration and water yield in northern China

There has been growing evidence that leaf area index (LAI) has changed in many parts of northern China, including northeast (NER), north (NR), northwest (NWR) and Huanghuaihai (HHH) regions. Changes in LAI can alter hydrologic cycle by regulating matter and energy cycles. However, the effects of the changes in LAI, particularly for afforestation/reforestation, on the hydrologic cycle have remained controversial. Better understanding on how the changes in LAI affect evapotranspiration (ET) and water yield has implications for ecohydrology, and regional and national afforestation/reforestation policies. In this study, we used satellite-derived LAI dataset and process-based ecosystem model (Boreal Ecosystem Production Simulator, BEPS) to examine how changes in LAI affect annual ET and water yield in northern China during the period 2000-2014. Firstly, LAI dataset was used to assess linear trends in LAI per pixel. Secondly, increasing and decreasing trends in LAI were removed on per-pixel basis by detrending LAI. Thirdly, we used the BEPS model along with the original and detrended LAI datasets to conduct two model simulations — one with original LAI (i.e., with increasing and decreasing trends in LAI) and one with detrended LAI (i.e., without trends in LAI). Finally, the two model simulations were conducted to assess the effects of increasing and decreasing LAI on terrestrial ET and water yield in northern China. Results showed significant trends in LAI for 20.2% of area of northern China, with increases and decreases in area accounted for 18.8% and 5.5%, respectively. Annual mean LAI averaged at the regional scale significantly increased for NR (0.001 4 a^-1, P < 0.05) and NWR (0.004 7 a^-1, P < 0.001). Temporal trends in LAI were not significant for NER and HHH. Two model simulations driven by original and detrended LAI, respectively, showed that the effects of the changes in LAI on terrestrial ET and water yield varied with spatial scale. At pixel scale, an increase in LAI increased annual ET but decreased water yield or weakened increase in water yield. However, a decrease in LAI decreased ET and increased water yield or weakened decrease in water yield. At regional scale, increase in LAI had positive effects on annual ET, but negative effects on water yield. Annual ET averaged at regional scale significantly increased in NR (3.5 mm·a^-1, P < 0.000 1) and NWR (2.1 mm·a^-1, P < 0.05). The difference in annual ET between the two simulations exhibited significant increase in the trends in all the 4 regions. Difference in water yield between the simulations with original and detrended LAI had decreasing trends for NWR (-0.95 mm·a^-1, P < 0.000 1), NER and NR (-0.38 mm·a^-1, P < 0.000 1) and for HHH (-0.11 mm·a^-1, P < 0.001). Future studies on the effects of the changes in LAI on hydrological cycle should account for feedbacks of the changes in ET and other biophysical properties (e.g., albedo) to the climate. Given the negative effects of increasing LAI on water yield and the water crisis in northern China, afforestation efforts should perhaps focus on southern China.