Creating a High Resolution Hydrologic Model for Simulating and Forecasting Floods in the Wabash River Basin
A research project funded by the Indiana Water Resources Research Center through the U.S. Geological Survey’s 104B annual base grants (section 104 of the Water Resources Research Act of 1984, as amended).
Start Date: 2019-05-31 End Date: 2020-05-30
Total Federal Funds: $14,990 Total Non-Federal Funds: $29,996
Indiana experiences major floods in different parts of the state almost every year from both excessive rain and rapid snowmelt. From February to April 2018, more than 20 counties in Indiana experienced extreme flooding and were declared disaster areas by the Governor’s office. Information related to rainfall and streamflow, two pieces of information critical for predicting floods, is not available for many areas in the state. Additionally, most flood inundation maps, created by the Federal Emergency Management Agency (FEMA), are static maps that cannot tell much about potential flood inundation during high storm events. Thus there is a critical need to create dynamic flood inundation information for most areas in Indiana during high storm events. To address this need, we created a hyper-resolution hydrologic model to provide street level flood information and inundation for all the areas within the Wabash River Basin.
Figure 1. Wabash River Basin study site. Hydro-logic model was created for area shown in blue to predict flood warning and help with long-term planning.
- Create a large-scale hydrologic and hydraulic model that can provide hydrologic fluxes and street level flooding information for any point in the entire Wabash River Basin. Various large datasets were used to acquire inputs related to topography, soil, land use information, precipitation, and hydrography for model development.
- Apply calibrated model to flood forecasting for the City of Indianapolis and Upper Wabash Basin. Two smaller models were created for sub-basins of the Wabash River Basin, the Upper Wabash Basin and the White River Basin encompassing the City of Indianapolis to efficiently analyze ways of improving model performance before scaling back up to the full-scale model.
Principal Investigator Dr. Venkatesh Merwade is a Professor in the Lyles School of Civil Engineering at Purdue University.
Major Conclusions & Significance
- By leveraging past and other ongoing modeling efforts in Minnesota and Texas, this project created a hydrologic model that can provide hydrologic fluxes for any area in the Wabash River Basin, which covers around 65% of the State of Indiana.
- This project also involved the development of automated tools for creating accurate geospatial descriptors such as river centerlines, banks, and river geometry.
- The resulting highly complex model has led to more research funding (approximately $632,000 from NSF) to make the availability of its tools and results to the wider research and general community, increasing the impact of this project.
Figure 2. Comparison of inundation extents estimated by the model and static FEMA maps. (a) 100-year FEMA flood extent map; (b) maximum flood extent during simulation; (c) flood inundation extent at 1.5 month time stamp; and (d) flood inundation extent at 2 month time stamp.
What Does This Mean For Indiana?
Indiana is one of the most flood-prone states in the country, experiencing major floods in different parts of the state almost every year from both excessive rain and rapid snowmelt. From February to April 2018, more than 20 counties in Indiana experienced extreme flooding and were declared disaster areas by the Governor’s Office. The impacts of climate change and land development and urbanization is expected to exacerbate both the severity and frequency of extreme flooding in the region. This project has developed a large-scale hydrologic and hydraulic model capable of providing detailed flood information across the Wabash River Basin. Comparison of model results with FEMA generated flood maps demonstrates the model’s ability to generate dynamic flood maps which can be used for both flood warning and long-term planning. The model performance has been validated against USGS gage measurements. The overall performance of the model is reasonable considering its scale. Additionally, two smaller models were created for the Upper Wabash Basin and the White River Basin encompassing the city of Indianapolis to explore techniques for further improving model performance. Flood simulations from the smaller models have demonstrated the effectiveness of the developed framework in creating locally relevant large-scale flood models capable of providing street level flooding information.
Training The Next Generation
One of the missions of the Indiana Water Resources Research Center, and all Water Centers, is to train the next generation of water scientists. This project successfully funded research for one Ph.D. student within Dr. Merwade’s lab.