FEMA's Tornado Damage Assessment Model and Lessons Learned from the 2019 Lee County, Alabama EF4 Tornado

A considerable challenge faced by the emergency management community at the start of a natural disaster is gaining situational awareness to begin response and recovery planning. The Federal Emergency Management Agency (FEMA) has looked to the creation of Python scripts and tools to help them increase the speed in which they are able to respond to disasters. Inside of FEMA’s Response Geospatial Office the Tornado Damage Assessment Model (TDAM) was developed in 2017 to serve as an automated way to rapidly estimate damages. The TDAM is a Python geoprocessing script tool that can predict damage to structures in near real-time following the availability of wind speed estimates and tornado path polygons within the National Weather Service’s (NWS) Damage Assessment Toolkit (DAT). The model incorporates nationwide CoreLogic parcel data and tornado path polygons from the DAT attributed with wind speed estimates and damage functions based on the Enhanced Fujita (EF) scale damage indicators. The data from CoreLogic includes land parcel outlines with information about land use and structure type and allows for damage functions to be constructed based on how increasing wind speeds affect varying structure types differently. This GIS-based model provides emergency responders the opportunity to cut lead time for preliminary damage assessments and improved situational awareness from 5 days down to 24-72 hours, or as soon as tornado path polygons are available within the DAT. The model was developed and tested on imagery-derived damage assessment data from 3 case studies (Joplin, MO 2011; Alabama 2011; Moore, OK 2013), resulting in 81-90% accuracy when the output damage assessments are compared to those derived manually from aerial optical imagery and field surveys. This year, the model was utilized as a situational awareness tool for informing FEMA’s response to the EF-4 tornado that ripped through Lee County, Alabama in March 2019. This proved to be an effective case study to continue to test model performance as well as reinvigorate collaboration between FEMA and the NWS offices and field survey teams. Here, we discuss outcomes and newly implemented changes to the model as a result of the lessons learned from this event. While examining this recent case study, we propose updates to the methodology and demonstrate how FEMA Response GIS is utilizing data provided by NWS storm survey teams to guide critical decisions about awarding financial assistance to tornado survivors.