Revealing the scale-specific influence of meteorological controls on soil water content in a karst depression using wavelet coherency

Temporal soil water content (SWC) patterns are affected by meteorological factors that operate at different intensities at different scales. Pearson's correlation analysis can only identify linear correlations at the measurement scale, therefore we examined the relationships between SWC and meteorological factors at different scales and occasions in a karst landscape using wavelet coherency analysis. SWC was measured at five soil depths using time domain reflectometry in two land uses over 242 d in a karst depression, southwest China. The results showed that the temporal SWC dynamics were influenced mainly by precipitation and land use. The temporal SWC patterns in farmland differed from those in grassland. The relationships between SWC and meteorological factors varied with scale, and a significant correlation was generally observed at large scales. Information on variation of scale for SWC improved the performance of soil moisture prediction. Soil depth had a greater impact on wavelet coherency at large scales than land use. Positive and negative correlations were detected between SWC and meteorological factors using wavelet coherency analysis. This study indicated that wavelet coherency analysis has potential to reveal the scale-specific dependence of SWC on controls.