Evidence of low watershed resilience across the Western United States

Vulnerable waters, including headwater streams and non-floodplain wetlands, are essential to watershed level resilience but notoriously difficult to measure over large spatial scales. Although individually small, vulnerable waters as a whole are integral in regulating hydrologic and biogeochemical processes. In the relatively small proportion of vulnerable waters that are continuously monitored, there are clear signs of declining water availability and resilience. However, the dispersed and remote nature of these waters makes continuous, landscape-scale monitoring impossible with traditional in situ methods, limiting our understanding of their condition. To address this gap, we produced a satellite-derived dataset of monthly ecologically-available water in vulnerable waters during the growing season from 2016 to 2024 across dryland ecoregions in the western United States. We then developed several indicators of conditions in vulnerable waters. Our results uncover varying levels of degradation in headwaters across the western U.S., and indicate that many watersheds may have very low resilience to climate or land-use shocks. In these watersheds, the functions provided by vulnerable waters, such as maintaining flow heterogeneity, sediment connectivity, water availability, and habitat heterogeneity, may be declining and threatened. We demonstrate the utility of the dataset we introduce in this study for identifying watersheds where functions have become limited and could be good targets for restoration efforts. Overall, we demonstrate that satellite-based measurements fill a major monitoring gap for vulnerable waters, and these continuous time-series measures across entire landscapes have the potential to transform our understanding of how vulnerable waters are changing over time. The metrics we calculated from the pixel-based maps are able to summarize important ecosystem function characteristics, and can be used to answer a variety of scientific questions and inform management decisions. Furthermore, our approach is completely open access and reproducible for future years and in other dryland biomes.