Temporal stability of soil moisture: a case study in a small watershed in the subtropical region of Brazil

Understanding the temporal behaviour of soil moisture in different environments is essential considering its importance in hydrological processes. Few studies have addressed its dynamics in a subtropical climate, in view of temperature variations throughout the year and well-distributed precipitation events which are characteristics that affect the behaviour of soil moisture over time. The temporal stability of soil moisture was investigated in two sets of data collected from a small watershed in southern Brazil. We used multiple linear regression (MLR) to evaluate whether points with greater temporal stability contributed as predictor variables of the models for estimating soil moisture. Soil moisture was monitored in the superficial (0-10 cm) and subsurface (10-20 cm) layers at 39 points with variations in soil characteristics and topography. To assess the temporal stability of soil moisture, two statistical techniques and MLR models were used to identify the topographic, climatological and soil property variables, as well as those that influenced the dynamics of the soil water content. The results showed that the factors of slope, terrain elevation, microporosity and evapotranspiration were determinants for the strong performance of the models, and that the inclusion of points with greater temporal stability helped to improve the estimate in certain datasets. Therefore, the results of this study indicate that the set of results that express topographic factors, land use and cover, along with the soil physical properties should be considered to characterise the temporal stability of soil moisture.