Leveraging catchment scale automated novel data collection infrastructure to advance urban hydrologic-hydraulic modeling

Lack of long-term hydrological observations in urban catchments presents a significant obstacle to advancing reliability and accuracy of urban flood models, as sensor-based depth, flow or velocity monitoring in stormwater drainage networks is costly and rarely available. Commonly available networked technologies e.g., traffic cameras and cell phones, collectively referred to as novel data sources, if properly leveraged, could provide a highquality, cost-effective source of urban hydrological observations. This proof-of-concept study utilizing experimental data collection infrastructure in Flagstaff, Arizona used remotely sensed flood-cameras based water depth time series and single point data, and citizens' contributed single point data for model parameterization. An approach to reduce the number of parameters while addressing spatial heterogeneity was applied before parameterization. The results demonstrate the reduced uncertainty and improved prediction accuracy from calibrated model even from a single event calibration, emphasizing the potential of novel data in advancing pluvial flood modeling in challenging urban contexts.