内蒙古草原的迅速退化正在引起严重的生态环境问题。尽管研究表明引起草地退化的主要因素有过渡放牧、农恳开发和气候变化，但对自然水文过程的改变原因和程度、触发草地退化的水文临界条件等问题还没有足够认识。本项目以内蒙古典型草原为研究对象通过野外试验、数据分析、遥感反演和构建模型相结合的方法，研究典型草原水文过程及其转换机理；并通过分析不同扰动和水土条件下草原水文过程揭示其水文扰动机理；并结合植被生理生态数据探讨草原植被对水文变化的响应；最后利用模型返演建立草原无扰动时的水文情况；将其作为基线来确定自然水文的改变程度，模拟分析在不同扰动情况下水文条件改变将如何影响草原退化，最终得到触发草地退化的关键水文过程及其临界条件和可逆性发生的机率。本研究对丰富和完善草地生态水文理论，揭示草地水文、植被、土壤动态响应关系有重要意义，也可为合理利用草地、维系草地可持续发展以及退化草地恢复提供理论支持和科学依据。

The rapid degradation and disappearance of native steppe grasslands in Inner Mongolia have caused serious environmental problems. Although previous studies identified that the main factors causing the degradation are overgrazing, cultivation, overdevelopment, and climate change, they have not examine how and to what extent the natural hydrology has been altered as well as threshold conditions triggering grassland degradation. This project will examine the dynamic interrelation of three important aspects (i.e., hydrology, soils, and grasses) of Inner Mongolia typical steppe by using remote sensing, field measurement, and state-of-the-art modeling to test a variety of competing hypotheses. We will conduct retrospective analyses to establish the "pristine" (i.e., least-disturbed) hydrologic conditions of steppe grasslands, which in turn will be used as the baseline for evaluating alterations of natural hydrology. In addition, we will develop an ecohydrologic model by linking variables representing these three aspects of steppe. The model will be used to simulate sensitivity of hydrologic alteration and/or topsoil erosion to grassland degradation, with different disturbances as a categorical variable. Further, we will determine threshold conditions triggering reversible and/or irreversible grassland degradation, and develop restoration scenarios. This project will advance the rationale and techniques of ecohydrology, and elucidate the interrelation among hydrology, soils, and vegetation. The project results will lay a solid scientific foundation for effectively utilizing and sustaining grasslands as well as restoring degraded ones.