Satellite-Monitoring Research Agenda for Urban Stormwater Infrastructure: Capabilities, Gaps, and a Community Benchmark Proposal

Abstract

Urban stormwater best management practices (BMPs) are a core component of urban-resilience portfolios worldwide, yet per-asset performance monitoring is limited and difficult to enumerate at the global scale. We argue that publicly available satellite Earth observation has reached the capability point at which this accountability gap can be closed at the portfolio scale and that the binding constraint is now community benchmark data rather than satellite data. We identify three operational capabilities available now: human-in-the-loop segmentation efficiently yields survey-grade asset footprints; the BMP type, historically the most difficult attribute to discriminate, is recoverable from learned multiphysical embeddings; and per-asset spectral departures beyond seasonality and regional weather are detectable. All three share the same underlying gap: a lack of community benchmark data that limits operationalization. The 36-site United States pilot supporting these claims is an existence proof of these capabilities rather than a deployment, and the identical workflow generalizes globally through ERA5-Land and globally available imagery. Closing the benchmark gap is a principal research agenda priority capable of converting present-day satellite capability into deployable accountability for infrastructure that today operates with little per-asset verification.