Research Team: Colin Phillips, Hongxing Liu, Rebecca Diehl, Erfan Goharian, Ibrahim Demir
Insitution: Utah State University, University of Alabama, University of Vermont, University of South Carolina, University of Iowa
Start Date: June 1, 2023 | End Date: May 31, 2025
Research Theme:
Longitudinal dynamics in the water conveyance potential of rivers are a function of changes in the channel dimensions, shape, roughness, and bed slope along river networks. Predictions of flow routing and flood inundation require sufficient information about channel and floodplain morphological (shape) and hydraulic (water flow) attributes and their streamflow dependence. However, representation of these attributes along stream networks remains elusive due to limitations in both observational data and computational resources. Emerging observational datasets (both field and satellite-based) and computational tools are transforming our ability to estimate these attributes across the United States. Exploiting these advances within large-scale operational hydrological prediction frameworks will require a new paradigm in the geospatial representation of morphological attributes. This project will build on ongoing CIROH projects and expertise to develop a new geospatial framework directly linked to the NOAA Office of Water Predictions (OWP) Next Generation (NextGen) streamflow and Flood Inundation Mapping (FIM) hydrofabrics (the geospatial representation of river systems). The project goals are to: (1) develop a flexible and scalable geospatial architecture to represent morphological attributes expected to impact flow routing and FIM prediction within the NextGen and FIM hydrofabric, (2) develop methods to estimate key morphological attributes from non-traditional data sources, and (3) evaluate improvements to streamflow routing and FIM model performance across a large range of channel settings. The first objective will be accomplished by combining observational data of various formats and features extracted from or predicted using available geospatial data. This objective directly supports NOAA’s long-term goal of developing a single channel geometry solution for NextGen. The second objective will leverage several methods including: channel and floodplain morphological attribute extraction from (i) high-resolution topo-bathymetric datasets and (ii) satellite multispectral imagery and laser altimetry datasets; and (iii) data-driven machine learning models to predict geomorphic attributes for reaches with insufficient data coverage or computational resources to apply more intensive methods. Key channel and floodplain properties will be identified, generated, and attributed within the hydrofabrics. Expected outcomes include automated data extraction tools, a suite of key morphological attributes calculated for a diverse set of channel reaches, and continuous, data-driven predictions of a subset of these attributes. This tiered approach will facilitate the development of a consistent, continuous morphological attribute dataset for efficient integration within the OWP hydrofabrics. In the third objective, the impact of the updated hydrofabric on hydrological and FIM prediction will be systematically evaluated for a range of regions and channel settings using the CIROH Community Water Model. We will quantify performance sensitivity and computational costs associated with improved representation of key morphological attributes. Performance improvements are expected to vary with channel and floodplain settings, highlighting future data and modeling needs. The project will result in improved representation of morphological attributes at selected locations within the hydrofabric that is optimized for efficient use within NextGen. The modular approach will provide a consistent framework for representing channels that is broadly applicable for large-scale hydrologic modeling (R2O) as well as the tools for OWP to generate more detailed, realistic attribute values where data and computational resources allow. Improved hydrological and flood model predictions have high potential to increase OWP forecast accuracy.