资源环境科技发展态势分析平台

Resilience of soil moisture regimes (SMRs) describes the stability of a particular SMR and its ability to withstand disturbances. This study analyzes the resilience of SMRs with quantifiable ecological (ECO-) and engineering (ENG-) metrics for a stochastic dynamic soil moisture system. The SMR is defined by the stationary state, described by a stationary probability distribution function (pdf), of the soil moisture dynamical system, and further classified into arid, semiarid, semiwet, and wet classes. Applying the stationary pdf of soil moisture dynamics derived by Rodriguez-Iturbe et al. (1999, ) and Laio, Porporato, Ridolfi, et al. (2001, ), the ENG- and ECO- resilience metrics of the various SMRs are quantified. We show that the recovery rate of soil moisture is a convex function of the expected soil moisture at the stationary state?the recovery rate reaches a minimum value at some intermediate soil moisture status. We also show that the maximum acceptable changes in the infiltration condition indicate the capacity of a system to avoid possible regime shifts. SMR shifts are characterized by phenomena of stagnation and hysteresis, which suggest two distinct thresholds for SMR shifts and their reversions. In particular, the semiwet SMR that is favorable to agriculture requires stricter infiltration conditions than other SMRs. This resilience analysis provides better understanding of how natural hydrological conditions control soil moisture, which helps provide guidance on maintaining SMRs suitable for agricultural activities and desertification prevention.