Workshop Title: FIMserv v.1.1: A tool for streamlining Flood Inundation Mapping (FIM) using the United States operational hydrological forecasting with Deep Learning Surrogacy
Day 2 Session 1 0(0:00AM)
Presenters:
Anupal Baruah, University of Alabama
Supath Dhital, University of Alabama
The National Oceanic and Atmospheric Administration (NOAA)-Office of Water Prediction’s (OWP) Operational HAND (Height above the nearest drainage) based FIM (Flood Inundation mapping) is a terrain-based flood inundation model that runs using the national water model (NWM) streamflow producing flood inundation maps at the watershed scale. FIMserv v1.0 is an automated, open-source, user-friendly, and cloud-enabled Python package to run this operational framework with some additional functionalities for producing FIMs across the country. The framework provides seamless integration of retrospective NWM, forecasted, recurrence streamflow and USGS gauge streamflow in flood inundation mapping. In FIMserv, we have integrated a deep learning-based surrogate model to enhance the FIM predictions by integrating the operational low-fidelity FIMs with other hydrological attributes. The workshop will provide hands-on training to the participants on generating FIM using the FIMserv v 1.1. Participants will learn how to generate FIMs and integrate the surrogate model to postprocess the FIM outcomes for specific AOI.
Learning Outcomes:
- Gain knowledge about the OWP HAND-FIM framework
- Learn how to access and use the FIMserv tool.
- Get familiar with the Deep Learning framework used in FIM.
Prerequisites:
- Prior knowledge: Brief understanding of flood inundation modeling, National Water model streamflow,
- Computing environment: Jupyter Notebook, Visual Studio Code, Google Collab
- Data: HUC8 id, USGS gauge location and NWM flow line ID (https://www.arcgis.com/apps/instant/basic/index.html?appid=88789b151b50430d8e840d573225b36b )