Temporal stability of surface soil moisture in a desert area of northwestern China

• Analyzed soil moisture temporal stability in a desert area.
• Spatial variation of soil moisture increased with increasing soil wetness.
• At locations with relatively high sand contents, soil moisture content tended to be more temporally stable.
• A representative site can be identified to accurately estimate field mean moisture content.
• Strong temporal stability due to bulk density, total porosity, soil texture and soil organic carbon.

SummaryThe patterns of soil moisture in arid environments are very important for the conservation and restoration of vegetation but have been rarely studied due to the difficulty of sampling in these environments. The characterization of the temporal stability of soil moisture and its statistical parameters is much needed. In this study, two sub-fields (southern and northern fields) with different topographies and soil properties in a region of the desert area in northwestern China were chosen to evaluate the temporal persistence of surface soil moisture, identify representative locations for reliably estimating the field mean moisture content, and investigate the factors influencing the temporal stability of soil moisture. Surface soil moisture was measured approximately every two weeks from April to October 2012 with a Theta Probe, for a total of 13 sampling campaigns. Soil moisture for the southern, finer-textured field was higher and more variable than for the northern, coarser-textured field. High Spearman’s rank correlation coefficients (p < 0.01) indicated a strong temporal stability of spatial patterns for both fields. Tests of relative difference indicated that soil moisture content was more stable under dry than under wet conditions. The representative locations identified for both fields accurately estimated the field mean moisture content, with coefficients of determination (R2) ranging from 0.981 to 0.998 and root mean square errors (RMSEk) ranging from 0.12% to 0.81% vol/vol. Among the methods for estimating the field mean moisture content using the representative locations, i.e. minimal SDRD (standard deviation of relative differences), ITS (index of temporal stability), MABE (mean absolute bias error), and RMSE (root mean square error), the minimal SDRD with a constant offset provided the best results. Correlation analysis revealed that (in descending order of importance) bulk density, total porosity, soil-particle composition, and soil organic carbon were the main factors influencing the temporal stability of soil moisture in this relatively flat region. Locations with relatively higher sand contents tended to have more pronounced temporal stabilities. This study contributes to our understanding of soil moisture patterns in arid environments, which have important implications for ground sampling design, hydrologic modeling, and sustainable land use.