Evaluation of FIM using Remote Sensing Workshop

Tools for Robust FIM Evaluation Using Remote Sensing Benchmarks

Day 4 Session 2 (11:00 AM MDT)

This workshop will present a new approach for evaluating predicted Flood Inundation Maps (FIM) using Remote Sensing-derived FIM (RS FIM) as the benchmark. Remote sensing-derived FIMs can be useful for evaluating the predictive skills of FIM models. There are, however, inherent biases and limitations in RS FIM which are often ignored by modelers, leading to improper use and erroneous evaluation results. Simply put, in many locations across the flooded domain, RS FIM can misclassify flooding conditions while the model prediction may be right. This issue is especially acute under dense vegetation and the built environment. During the workshop, we will provide a short overview and examples of the issues with RS FIM use in model evaluation. We will then provide the background for a new strategy for more robust FIM evaluation using RS FIM. Building on the ‘Remote Sensing Flood Inundation Mapping and Enhancement with High-Resolution DEM’ workshop, we will use our RS FIM to run the RS FIM enhancement and evaluation tools