Improvement of lake and reservoir parameterization in the NOAA National water model

The development of operational, region-wide hydrologic forecasting systems requires a large-scale modeling approach that includes various physical components and a representation of their interactions. Such modeling systems rely on a significant amount of pre-processing efforts in terms of model parameterization. The NOAA National Water Model (NWM) is a recent example of a large-scale hydrologic modeling system with the primary goal of forecasting streamflow and other hydrologic states across the river basins of the Contiguous United States (CONUS). The NWM version 2.1 includes 5,781 lakes/reservoirs (hereafter, waterbodies), which represent a variety of geophysical attributes and outflow conditions (managed versus unmanaged). Major data sets that currently support features of the NWM rivers and waterbodies are based on the National Hydrography Dataset Plus (NHD+), and from data sets managed by the United States Army Corps of Engineers (USACE) and the U.S. River Forecast Centers (RFC). However, the assessment of the NWM's performance of discharge below managed reservoirs underscores the need to revise the waterbody geometry and discharge characteristics. In this study, we investigate the sensitivity of waterbody outflows and the water balance to different geometric assumptions such as shape attributes and head-area-volume (h-A-V) relationships. We also investigate the sensitivity of discharge characteristics of these waterbodies (e.g., weir and orifice parameters) and their impact on the streamflow prediction in the NWM. We evaluate various aforementioned model parameters and configurations on waterbody discharge against streamflow observations from a set of USGS stations.