Wetlandscape Configuration and Structure as Predictors of Watershed Hydrologic Signatures

Traditional hydrologic models often treat wetlands as an aggregated unit, failing to capture dynamic wetlandscape processes, including seasonal expansion-contraction cycles and evolving connectivity patterns that fundamentally alter watershed flow generation. We hypothesized that the inclusion of wetlandscape patterns and behavior can improve prediction of watershed hydrology. Here, we link wetlandscape characteristics to hydrologic signatures across 200 gaged watersheds in the Prairie Pothole region in the US. Static and dynamic landscape metrics capturing wetland position, configuration and variability were determined using the Dynamic Surface Water eXtents from Harmonized Landsat Sentinel-2 (DSWx-HLS) product and the National Wetlands Inventory while hydrologic signatures were derived from daily discharge gages over twenty years of observation in watersheds with a range of wetland densities. Regression models explained hydrologic signature magnitude and variability using and wetlandscape metrics together with climate, topography, land cover, as predictors. The addition of wetlandscape configuration explained significant additional variance beyond traditional watershed characteristics for multiple signatures. Variable importance analysis revealed wetland spatial patterns ranked among top predictors for six of eight signatures examined. These findings demonstrate that incorporating spatially-explicit wetlandscape dynamics substantially improves hydrologic prediction capabilities across multiple temporal scale