Enhancing River Channel Dimension Estimation: A Machine Learning Approach Leveraging the National Water Model, Hydrographic Networks, and Landscape Characteristics

• Published in: Journal of geophysical research. Machine learning and computation Vol. 1; no. 4
• Main Authors: Modaresi Rad, Arash, Johnson, J. Michael, Ghahremani, Zahra, Coll, James, Frazier, Nels
• Format: Journal Article
• Language: English
• Published: Wiley 01.12.2024
• Subjects: at a station hydraulic geometry; hydrofabric; machine learning; National water model; river characteristics; river dimension
• ISSN: 2993-5210; 2993-5210
• DOI: 10.1029/2024JH000173

Knowledge of bankfull hydraulic geometry represents an essential requirement for various applications, including accurate flood prediction, hydrological routing, river behavior analysis, river management and engineering practices, water resource management, and beyond. Our work builds upon an extensive body of literature about estimating bankfull top‐width and depth at ungauged locations to enhance the understanding of observable factors that affect these parameters. Using more than 200,000 USGS Acoustic Doppler Current Profiler (ADCP) records, we developed a method employing machine learning (ML) using discharge estimates and landscape characteristics from sources, including the National Water Model (NWM), the National Hydrologic Geospatial Fabric network (NHGF), the EPA stream characteristic data set (StreamCat), and an array of satellite and reanalysis data products. Our method achieved log‐transformed R2 = 0.8 predicting bankfull depth (R2 = 0.77 for in‐channel conditions) and R2 = 0.76 predicting bankfull top‐width (R2 = 0.66 for in‐channel conditions) in the testing data set. The depth and width predictions showed lowest skill in mountainous and plateau regions. Our analysis demonstrates the benefit of data‐driven modeling in contrast to other global scaling‐based or regional statistical methods. In summary, our study illustrates how top‐width and depth can be better predicted using ML, reanalysis streamflow simulations, hydrographic networks, and summarized geospatial data. Plain Language Summary Accurately estimating (or generalizing) key characteristics of river channels, such as their top‐width and depth, is valuable for tasks like predicting water flow, modeling water‐related processes, and mapping flooded areas. Our research builds on existing studies that focus on estimating these important channel characteristics and aims to further develop knowledge and skills in predicting these channel characteristics. In this work, we use over 200,000 historical measurements of channel top‐width and depth to develop a machine learning (ML) model to estimate channel top‐width and depth. The model uses widely available information from the National Water Model (NWM) discharge and other data sets that represent land surface characteristics, climate, hydrographic connectivity, and human‐related structures. The developed model performs well compared to other global, regional, and ML‐based methods in the literature within the Continental United States. Validation of the models across different regions indicated better performance in flatter regions and lower performance in steeper areas. In conclusion, the study highlights the advantages of using ML techniques to estimate channel geometry more accurately, paving the way for improved predictions in unmeasured channels. A hydrographic network‐based machine learning approach for estimating river channel width and depth, integrating landscape, climate, and catchment attributes.