Statistical analysis of the temporal stability of soil moisture in three desert regions of northwestern China

Soil moisture and its variations are key factors for understanding hydrological processes, which are characterized by a high temporal variability at different scales. The study was conducted at three field stations in the desert regions of northwestern China, where soil moisture measurements with gravimetric method were used to characterize the temporal stability of soil moisture using various statistical parameters and an index of temporal stability (ITS). The soils are a gray–brown desert soil at the Linze station, an aeolian sandy soil at the Fukang station, and a brown desert soil at the Cele station. Soil textures are accordingly sandy loam at Linze and Cele, and loamy sand at Fukang. The dynamic variation in soil moisture depends strongly on the rainfall pattern (amount and frequency) in these desert ecosystems. Soil moisture content is low and significantly different among the three desert ecosystems, with the maximum at the Linze station (6.61 ± 2.08 %), followed by the Cele (4.83 ± 0.81 %) and Fukang (3.46 ± 0.47 %) stations. The temporal pattern exhibits high variability because soil moisture is characterized by low temporal stability and a high coefficient of variation (CV). The standard deviation, CV, and ITS increase significantly with increasing soil moisture. Soil moisture displays a skewed frequency distribution that follows a logarithmic function at lower soil moisture but a log-normal distribution at higher values.