IDRT has developed a new web tool to show the annual likelihood of flooding based on historic flood damages. To accomplish this, researchers at the Institute for a Disaster Resilient Texas at Texas A&M University at Galveston used a machine learning algorithm trained on nearly 40 years of insured loss data from FEMA’s National Flood Insurance Program. The output of this model can be thought of as a “damage plain” because it illustrates the likelihood that a home will experience a structurally damaging flood event any given year.