The counteracting effects of large-scale vegetation restoration and increased precipitation on drought in the Huang-Huai-Hai-Yangtze River Basin
Mengyang Ma,Qingming Wang,Yong Zhao,Rong Liu,Enchong Li
DOI: https://doi.org/10.1016/j.jhydrol.2023.129733
IF: 6.4
2023-06-03
Journal of Hydrology
Abstract:Quantifying the causes of drought trends is critical for disaster prevention and water resource management. In recent years, not only climatic variables, such as temperature and precipitation (PRE), have changed, but also environmental factors, including vegetation restoration, increased CO 2 concentration, and land cover conversion. However, the role of diverse factors, particularly large-scale vegetation restoration, in influencing drought trends is still indeterminate. In this study, a Ball-Berry canopy conductance model was incorporated into the Shuttleworth–Wallace (S-W) model to estimate the potential evapotranspiration (PET) in the Huang-Huai-Hai-Yangtze River basin from 2001 to 2019. The standardized precipitation evapotranspiration index (SPEI), which considers both water supply (from PRE) and demand (from PET), was adopted to identify drought conditions. Given the complex non-linear relationships in the model, an interpretable machine learning model was applied to the attribution of PET and SPEI trends. The results were as follows: (1) The PRE and PET showed significant upward trends, whereas the change rate of PET (60 mm/10a, p<0.01) was higher than that of PET (44 mm/10a, p 0.1). Increasing PET dominated SPEI trends with a contribution of -0.28/10a (p 0.1); (3) The interpretable machine learning model performed better than multiple linear regression in attributing PET trends; (4) Large-scale vegetation restoration (the increased leaf area index, LAI) and wind speed, were the two dominant factors influencing PET trends with contributions of 32.4 mm/10a (32%) and 19.6 mm/10a (20%), respectively; (5) In terms of the trends in drought across the entire region, the increase in LAI almost offset the positive contributions of PRE to SPEI variations, hence, combined with the influence of the other factors, drought was non-significantly accelerated.
geosciences, multidisciplinary,water resources,engineering, civil