Average moisture saturation effects on temporal stability of soil water spatial distribution at field scale

► Surface water content data were measured by TDR in an experimental field. ► They were analyzed by spatial and temporal statistical techniques. ► Water content showed reversed spatio-temporal patterns in drier and wetter stages. ► Analyzing temporal stability should account for dependence on soil saturation degree. Surface water content data measured by TDR in the tilled zone of an experimental field were analyzed by spatial and temporal statistical techniques. First, spatial autocorrelation was explored by using isotropic variograms. The temporal persistence of the spatial pattern was verified by calculating Pearson's correlation coefficients. Second, a temporal analysis technique was tested, aiming to identify sites with time-invariant statistical properties of the probability density function. These locations would provide water content measurements consistent with the mean or given percentile values for the whole field. The results showed two different behaviors of the surface water content, the first for a drier stage with no or little rainfall, the second arising after more significant rainfall. In neither stage was a significant spatial dependence found. However, in the drier stage, the experimental semivariance calculated for each spatial lag was lower than in the wetter stage. Also, a reverse pattern of the water content was observed in the two stages, with drier zones becoming wetter and vice versa. For each stage, a significant temporal stability of the water content pattern was proved, which was confirmed by calculating Pearson's correlation coefficients. The percentile at each measurement location varied over time, with more significant changes between the drier and wetter stages. Nevertheless, for a few sites the percentile was quite stable during both the dry and wet stages. Site 34 was steadily in the range 0.4–0.6 and was identified as being representative of the average water content in the investigated field for both dry and wet seasons. This was confirmed by calculating Spearman's rank coefficients.