Localized and severe storms can cause citywide flooding, leading drainage systems to surcharge and overflow to nearby water courses. Urban catchments feature high degrees of imperviousness and heterogeneity, often resulting in highly nonlinear hydrologic responses with shorter time of concentration, lag times, and sharper peak flows. Additionally, due to population and economic growth, urban drainage systems have attempted to evolve to more efficiently drain surface waters and reduce vulnerabilities. A critical outcome of this evolution is the need for finer spatio-temporal resolution rainfall measurements and hydrological modeling. As the major driving mechanism, the spatio-temporal variability in rainfall is acknowledged as a key source of uncertainty for urban hydrological modeling. The objective of this research is to revisit the impact of the temporal and spatial resolution of rainfall measurements on urban hydrological applications. We first provide a quantitative analysis of the spatiotemporal structure and variability of rainfall using both a 9-member hourly rain gauge network spaced ~10 km apart and a single WSR-88D dual-polarimetric weather radar with precipitation resolved every 5 minutes at ~500 m. Precipitation data from each observing system extracted at different time steps is aggregated within urban catchments and compared for three typical intense storms over a set of urban catchments located in Chicago Metropolitan area. Then the rain-runoff dynamics for 9 geographically-diverse (relative to the underneath sewer system) and differently-sized catchments are examined utilizing MetroFlow – a coupled hydrologic and hydraulic modeling system. Finally, city-wide flooding risks are simulated by routing the predicted surface runoff through the as-built sewer system. Additional mitigating storage capacity is also considered by numerical modeling the deep tunnel and reservoir in construction. The sensitivity of urban flood variables (i.e., mean and peak depth as well as duration) to rainfall spatiotemporal resolution is analyzed. Our results complement and advance the limited literature attempting to resolve the ideal resolution of rainfall data relevant for urban hydrology and stormwater management.

Assessing the extent and impact of past extreme weather events within cities can help identify vulnerabilities, map potential solutions, and prevent future calamities. Modern urban environments are particularly vulnerable to hydrological extremes due to high population densities, expansive impermeable surfaces, more intense precipitation extremes, and infrastructure designed for a now obsolete climate. Intense rainfall in urban environments can lead to impacts ranging from nuisance flooding to overloading of sewage and drainage systems to neighborhood inundation. In the flood-prone city of Chicago, storm waters are contained in a network of tunnels and reservoirs until treated and released to the waterways. Management decisions for a 600 km^2 metropolitan area are made based on precipitation data collected at just 9 gauge sites. Here, we combine high-resolution radar-derived precipitation data with urban-scale hydrological models to improve our understanding of water flow, advance stormwater management practices, and potentially mitigate flood risks. Proximity of the NEXRAD system to Chicago allows us to improve the spatial resolution of rainfall estimates to 500m, which will be used to produce neighborhood-scale rainfall hindcasts. Different dual-polarimetric radar-rainfall retrieval methods, e.g., rainfall from reflectivity, attenuation, specific differential phase, and differential reflectivity will be examined to determine the most accurate representation of rainfall estimates. This suite of rainfall estimates will be used to derive catchment-level precipitation, and serve as input in a coupled hydrological-hydraulic MetroFlow model. To verify the utility of our radar precipitation data, we examine an April 2013 event that delivered a record-breaking 7 inches of rain in 2 days in some areas. We compare our highly-resolved precipitation-driven hydrological model predictions with those made using the 9 gauge stations. This research is conducted under the premise that hydrological extremes are expected to be exacerbated by climate change. Understanding drivers of urban flooding using high-resolution precipitation data and models can be used to improve resiliency-focused infrastructure design in Chicago neighborhoods.