古尔班通古特沙漠南缘固定沙丘坡面土壤水分动态随机模拟

为揭示土壤水分动态对干旱沙区坡面尺度水文和生态过程的影响机理,基于Rodriguez-Iturbe土壤水分动态随机模型,通过修正模型输入项参数,获取干旱沙区坡面土壤水分动态随机模型,并利用2015—2016年古尔班通古特沙漠南缘固定沙丘坡面生长季根系层土壤水分观测数据及2000—2016年日降水资料,分析研究区生长季坡面土壤水分动态与土壤水分概率密度函数特征,探讨该坡面随机模型参数敏感性及应用可行性.结果表明: 研究区生长季降水分布极不规则,以<5 mm降水为主,降水量主要源于>10 mm降水,降水间隔期以<10 d为主,且东坡次降水量略大于西坡;东、西坡面生长季土壤水分年际变化与降雨分布基本一致,西坡根系层土壤含水率略低于东坡,且各坡面生长季根系层土壤水分均服从正态分布;该坡面随机模型模拟得到的不同坡面生长季根系层土壤水分概率密度函数的曲线特征(峰值及其位置、95%置信区间)与观测结果基本一致,一致性指数(CM)均大于0.5,说明该坡面随机模型能对研究区坡面土壤水分概率密度函数进行高效模拟,具有较好的适用性,可为干旱沙区植被生存或恢复研究及实施荒漠化防治生态建设工程提供理论基础和科学方法.

Hillslope-scale stochastic simulation of soil moisture dynamics in fixed dunes at the southern edge of Gurbantunggut Desert,China.

To examine the effects of soil moisture dynamics on hydrological and ecological processes at the hillslope scale in the desert region, a hillslope-scale soil moisture dynamic stochastic model in arid sandy based on Rodriguez-Iturbe soil moisture dynamic stochastic model (RI model) was developed by modifying the input parameters of the model. Using the continuously monitored data of soil moisture in root zone during growing season from 2015 to 2016 and daily precipitation from 2000 to 2016 in a fixed dune at the southern edge of Gurbantunggut Desert, we analyzed soil moisture dynamics and probability density function of soil moisture on slopes in growing season and discussed the sensitivity of parameters of the stochastic model and its application feasibility. The results showed that the seasonal distribution of precipitation was irregular. The total number of annual precipitation was mainly derived from the precipitation of >10 mm. The precipitation frequency and interval were dominated by <5 mm and <10 d, respectively. The precipitation in the eastern slope was slightly larger than that in the western slope. The interannual variability of soil moisture in the eastern and western slopes in the growing season was basically consistent with the rainfall distribution. The soil moisture in the root zone in the western slope was slightly lower than that in the eas-tern slope, and the soil water content in different slope positions followed a normal distribution. The curve characteristics of the soil moisture probability density function (the curve peak value, the position of the peak, the confidence interval of 95%) in the root zone in the different slopes in growing season simulated by the hillslope-scale model were consistent with observed values, with the consistency measure (CM) being higher than 0.5. Such a result indicated that the hillslope-scale stochastic model could be used for simulating the soil moisture probability density function in diffe-rent slope positions in the study area, with a good applicability. This model could provide theoretical basis and scientific methods for vegetation survival or restoration and the implementation of ecological construction project of desertification control in arid sandy area.