A New Global River Network Database for Macroscale Hydrologic Modeling
River networks at coarse resolutions are critical to providing information for macroscale hydrologic models to accurately represent fresh water redistribution over spatial and temporal domains. The newly developed Dominant River Tracing (DRT) algorithm advances all previous existing river network delineation methods by preserving the baseline fine-scale hydrography information in a comprehensive manner and significantly reducing the distortions.
Researchers, including those at DOE’s Pacific Northwest National Laboratory, developed a new global dataset based on this start-of-the-art algorithm and the most recent global fine-scale hydrography dataset, HydroSHEDS. In this study, the DRT algorithm was modified for improved computational efficiency. More importantly, the new river network dataset was evaluated at multiple coarse spatial resolutions against the original baseline high-resolution hydrography. It is found that the performance of the new dataset has systematically improved over previous datasets.
The improved baseline hydrography inputs, together with the improved computational efficiency, enable greater accuracy in the derived DRT upscaled river networks, which can facilitate more accurate regional and macroscale hydrological model simulations that utilize these data.
River networks at coarse resolutions are critical to providing information for macroscale hydrologic models to accurately represent fresh water redistribution over spatial and temporal domains. The newly developed Dominant River Tracing algorithm advances all previous existing river network delineation methods by preserving the baseline fine-scale hydrography information in a comprehensive manner and significantly reducing the distortions. Researchers, including those at DOE’s Pacific Northwest National Laboratory, developed a new global dataset based on this start-of-the-art algorithm and the most recent global fine-scale hydrography dataset, HydroSHEDS.
In this study, the DRT algorithm was modified for improved computational efficiency. More importantly, the new river network dataset was evaluated at multiple coarse spatial resolutions against the original baseline high resolution hydrography. It is found that the performance of the new dataset has systematically improved over previous datasets. The improved baseline hydrography inputs, together with the improved computational efficiency, enable greater accuracy in the derived DRT upscaled river networks, which can facilitate more accurate regional and macroscale hydrological model simulations that utilize these data.
This work was conducted at the University of Montana (UMT) and Earth System Science Interdisciplinary Center (ESSIC), University of Maryland, with financial support from the Gordon and Betty Moore Foundation, the NASA Applied Sciences Program (Michael Goodman), and the integrated Earth System Modeling (iESM) project funded by the DOE Earth System Modeling Program.