Improving Urban Flood Modeling with the Integration of Novel Data Sources
Abstract
Flooding is the most costly natural hazard globally. The vast majority of flood risks to life and property are concentrated in our cities, yet, the observation and forecasting of streamflow and floods in United States is mostly focused on relatively large rivers and, rarely cover fine scale urban streams and washes. With urban flood risks projected to rise with increasing extreme precipitation events and further land use change, improving our ability to sense, understand and predict urban flooding is critical. Hydrological and hydraulic modeling can increase understanding and predict flood extent based on precipitation observations or forecasts if sufficient data is available for model development and testing. The current dearth of flow measurement in the urban environment preclude model calibration and validation. Here we address this challenge by pairing road level water depth measurements, derived from images taken from our team's flood cameras, with a high resolution model of urban hydrology and hydraulics developed in EPA SWMM for the city of Phoenix, AZ. In this study, we instrumented known flooding hot spots with cameras to remotely estimate water level on street during flood events, to calibrate the model. We then developed a two-dimensional storm water model in EPA - SWMM using infrastructure design data, USGS 3DEP Lidar derived digital elevation model, and addition topographic data from the Maricopa County Flood Control District. Here we present the data integration process and preliminary results from our first field season.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H31I1826S
- Keywords:
-
- 1805 Computational hydrology;
- HYDROLOGY;
- 1875 Vadose zone;
- HYDROLOGY;
- 1879 Watershed;
- HYDROLOGY;
- 1942 Machine learning;
- INFORMATICS