Temporally stable patterns but seasonal dependent controls of soil water content: Evidence from wavelet analyses

Abstract

Scale- and location-dependent relationships between soil water content (SWC) and individual environmental factors have been widely explored. SWC is controlled by multiple factors concurrently; however, the multivariate relationship is rarely explored at different scales and locations. Multivariate controls of SWC at different scales and locations in two seasons within a hummocky landscape of North America were identified using bivariate wavelet coherency and multiple wavelet coherence. Results showed that depth to CaCO₃ layer which was correlated with elevation over all locations at scales of 36–144 m and cos(aspect) provided the best individual factor for explaining SWC variations in the spring (May 2) and summer (August 23), respectively. Although spatial patterns of SWC were temporally stable, different topographic indices affected spatial distribution of SWC in different seasons (elevation in spring and aspect in summer) due to different dominating hydrological processes. These varying hydrological processes also resulted in the distinct role of soil organic carbon (SOC) content in different seasons: a positive correlation in the spring and a negative correlation in the summer. Multiple wavelet coherence identified a combination of depth to CaCO₃ layer and SOC in spring and a combination of cos(aspect) and SOC in summer that controlled SWC at different scales and locations, respectively. This indicated a combined effect of soil and topographic properties on SWC distribution and a clear need for these two factors in developing scale-dependent prediction of SWC in the hummocky landscape of North America.