College of Agriculture and Natural Resources
Center for Land Use Education and Research
NASA DEVELOP

NASA DEVELOP Summer 2006 Project

Modeling Coastal Inundation Resulting from Tropical Storm Surge

 

Results

The results from the accuracy assessment proved that the LIDAR data was the most accurate DEM used. With only a RMSE of 2.81ft for the LIDAR 10ft, it seemed the most suitable for generating flood maps. The SRTM and NED data seemed unsuitable for flood prediction because they both had a RSME over 10ft. With that large of an error predicting a flood of 5-10ft is unrealiable.

Figure 1. The graph of the accuracy assesment for the six different DEMs.

The use of different elevation data in the creation of the flood maps has a profound effect on the result. By using different elevation datasets each flood map encompassed vastly different areas. These different areas covered different residential zones, which in turn affected the number of buildings inundated. This has startling implications on emergency management. The use of an inaccurate elevation dataset could produce very unreliable flood maps. This project proved the need for the use of the most accurate elevation data available when creating flood maps for emergency planners. Below are flood maps generated for a Category 3 hurricane moving in a NNE direction. Notice how the area flooded is different when using different DEMs.


Figure 2.  Three flood maps created by using NED data, LIDAR data, and SRTM data.  The red circles indicate areas of disagreement between maps.

In order to provide the most accurate data for emergency management planners and evacuation crews having the most accurate flood maps is essential. Since each DEM creates different flood maps it is important to use the most accurate DEM. For these reasons the NASA DEVELOP team concluded that the flood maps generated using the LIDAR 10ft data would be the most effective to minimize damages associated with tropical storms. Currently the 2004 FEMA flood inundation maps created for the coastal towns of Connecticut did not use LIDAR data. This brings into question the accuracy of the maps currently used my emergency planners. The NASA DEVELOP team strongly recommends upgrading the states flood maps using LIDAR data based on the results of this study. Hopefully the Connecticut coastline will be more prepared for severe hurricanes in the future.

Back to Top