Merging Imagery and Models for River Current Prediction

摘要

To meet the challenge of operating in river environments with denied access and to improve the riverine intelligence available to the warfighter, advanced high resolution river circulation models are combined with remote sensing feature extraction algorithms to produce a predictive capability for currents and water levels in rivers where a priori knowledge of the river environment is limited. A River Simulation Tool (RST) is developed to facilitate the rapid configuration of a river model. River geometry is extracted from the automated processing of available imagery while minimal user input is collected to complete the parameter and forcing specifications necessary to configure a river model. Contingencies within the RST accommodate missing data such as a lack of water depth information and allow for ensemble computations. Successful application of the RST to river environments is demonstrated for the Snohomish River. WA. Modeled currents compare favorably to in-situ currents reinforcing the value of the developed approach.