As a common measure for grassland restoration, the fence for prohibition of grazing is usually applied to the steppe in northern China. A study in Hulunbuir grassland showed the average litter mass in the steppes that enclosed for ten years can reach 70-550g/m². The excessive accumulation of litter caused by long-term grazing prohibition can intercept more precipitation, which will reduce the water available to grassland plants and hinder the further recovery of plant communities. Therefore, the establishment of a reasonable litter interception model can help estimating the interception of steppes with different litter accumulation and provide some theoretical support for the formulation of fencing policies. Based on the model of litter interception rate change, we established a dynamic interception model of litter: $I=I_m\left(1-e^{-kt}\right)$ ($t$: time;$I$: dynamic interception; $I_m$: maximum interception; $k$: exponential coefficient of litter interception). After testing by an indoor simulated rainfall experiment, we found that this model has high accuracy ($R^2:$0.908 – 0.992). The litter mass (LM) determines the maximum interception, and the rainfall intensity (RI) determines the speed of interception which reaches its maximum value. After the conversion of coefficients, we established a prediction model of litter interception based on LM and RI: $I=\left(0.003LM+0.44\right)\left[1-e^{-\left(0.33RI+3.76\right)t}\right],R^2=0.86,P<0.001$. The easily obtainable characteristics of LM and RI data make this model easy to apply.

作为草原恢复的常用措施，我国北方草原普遍采用禁牧围栏。呼伦贝尔草原研究表明，封闭十年的草原平均凋落物量可达70-550g/m ². 长期禁牧造成的枯枝落叶堆积过多，可以截留更多的降水，这将减少草原植物可利用的水分，阻碍植物群落的进一步恢复。因此，建立合理的凋落物截留模型有助于估算不同凋落物堆积的草原截留量，为围栏政策的制定提供一定的理论支持。基于垃圾拦截率变化模型，我们建立了垃圾的动态拦截模型：$\mathrm{一世}={\mathrm{一世}}_{米}\left(1-e^{-ķ吨}\right)$ ($吨$： 时间;$\mathrm{一世}$：动态拦截； ${\mathrm{一世}}_{米}$：最大拦截； $ķ$：垃圾拦截的指数系数）。经过室内模拟降雨实验测试，我们发现该模型具有较高的精度（$R^2：$0.908 - 0.992）。凋落物质量（LM）决定了最大截获量，降雨强度（RI）决定了达到最大值的截获速度。系数转换后，我们建立了基于LM和RI的垃圾拦截预测模型：$\mathrm{一世}=\left(0.003\mathrm{大号}米+0.44\right)\left[1-e^{-\left(0.33R\mathrm{一世}+3.76\right)吨}\right],R^2=0.86,磷<0.001$. LM 和 RI 数据易于获取的特性使该模型易于应用。