Introducing Ripple 1D: Software for extending the life and applicability of HEC-RAS models for operational flood forecasting

The Federal Emergency Management Agency (FEMA) has developed a large inventory of flood models created in support of the National Flood Insurance Program (NFIP). The majority of the hydraulic models in this inventory were developed using HEC-RAS software, provided by the US Army Corps of Engineers, and used to simulate frequency-based flow scenarios (i.e. 1% annual exceedance probability) to publish regulatory flood maps. These archived models were identified by the National Weather Service (NWS) Office of Water Prediction (OWP) as potential data sources that could be used to enhance and validate the baseline Height Above Nearest Drainage Mapping capabilities. Building on OWP’s existing software RAS2FIM, that repurposes HEC-RAS models for applications of the National Water Model, Ripple 1D is a new software that expands the potential for migrating FEMA models for use in operational forecasting. Features of this software include incorporating backwater on a reach basis, improving accuracy of the network hydrologic system, and minimizing discontinuities at confluences through the incorporation of know water surface elevations. Additionally, Ripple 1D enables the integration of existing hydraulic structures in reach based reach-based models and provides conflation metrics for translating HEC-RAS cross sections to different hydro-fabrics. This talk will provide an overview of the technical implementation of the hydraulic modelling challenges and processes as well as mechanisms put in place to enable scaling on the Windows OS, which is required for the HEC-RAS software. Backed by a lightweight scheduling and tasking system, commands are passed to Ripple 1D via a RESTful API. With the goal of supporting interoperable frameworks, the API was designed and built to adhere as closely as possible to the OGC Process API specification. Similarly, the metadata and input schemas were designed using another interoperable feature, the SpatioTemporal Asset Catalog (STAC), which provides a consistent and versatile standard for human readable data implemented in the production workflow. Examples of dynamically generated flood maps developed using FEMA source models will be reviewed with a discussion of challenges and limitations.