Improving Laio’s model for investigating soil water response to vegetation restoration in the middle Yellow River Basin, China

Abstract

Over the last few decades, vegetation–soil water interactions in the semiarid middle Yellow River basin have been a topic of interest for many researchers, especially since the implementation of the Grain-to-Green Project (GTGP) in 1999. The probabilistic states of soil water and water stress on vegetation development provide interactions from a perspective that is different from process-based modelling approaches. The Laio’s model, a widely used probabilistic soil water model was applied in the study. To include the infiltration-excess runoff, co-exist with the saturation-excess runoff, or even be dominant in the study region, a parameter Δ2 was incorporated to reflect mean surface runoff of a rainfall event under current theoretical framework of Laio’s model. The modified model was evaluated against the observations, and the effectiveness of the modification was proved in the study region. Based on the modified model, the soil water response to vegetation restoration was studied with remote-sensing based land use and leaf area index in the 2000 s and 2010 s. Vegetation was restored in most parts of the region, resulting in soil water reduction in approximately 83 % of the area, and the soil water stress increased by 49 % on average. The reduction in soil water and increase in soil water stress were more serious in areas that had converted vegetation and loess soil than those had non-converted vegetation and non-loess soil. These findings might imply that vegetation greening will reach an upper limit due to the negative soil water balance and increased water stress on plant growth, which is meaningful for a better understanding of vegetation–soil water interactions and ecosystem management in semi-arid regions.