The value of increased spatial resolution for inundated area mapping in the Western United States

Monitoring surface water dynamics is critical for hydrologic assessment in drought prone and climate sensitive regions. This study evaluates differences between two widely used satellite-derived inundation products: the Landsat-based Global Surface Water dataset (30 m) with the Sentinel-based Dynamic World Global Surface Water dataset (10 m). Using the Colorado River Basin as a case study, basin-scale and stream-adjacent inundation patterns are compared to assess how differences in spatial resolution, observed frequency, and classification methodology influence mapped surface water extent. Results indicate that the Sentinel-derived product detects more than twice as many low-order streams segments than the Landsat-based product, with the largest discrepancies occurring in snowmelt-dominated headwater regions. While agreement between products is high for large, stable water bodies, divergence increases in narrow or fragmented channels. These findings demonstrate that commonly used inundation products can yield systematically different representations of stream connectivity and cumulative water extent. The results highlight the importance of understanding product-specific characteristics when interpreting long-term surface water patterns, particularly in semi-arid regions where small or ephemeral features are hydrologically significant.