Impacts of high-frequency radar-derived wind observations on COAMPS® forecasts

Abstract:

The U.S. Navy requires accurate characterization and prediction of meteorological and oceanographic conditions for effective naval operations. The assimilation of new types of environmental observations have the potential to improve the numerical weather prediction (NWP) models relied on by the Navy. For decades, high-frequency (HF) radar has been used to remotely measure ocean currents and wave characteristics up to 200 km from shore. Recently, a new method has been developed by SRI International to derive near-surface winds from HF radar Doppler spectra data with the adjoint of the HF radar model and the Simulating WAves Nearshore (SWAN) model. The potential benefit of this new wind dataset for data assimilation and improved NWP is an exciting advance in remote sensing of coastal winds and expands the benefit of the existing HF radar network beyond its intended use.

A limited dataset of HF radar-derived wind observations has been generated using this new technique during the CASPER-West field experiment in October 2017 near Santa Barbara, California. In this study, we test if these novel wind observations improve wind forecasts in the littoral zone between the Channel Islands and the Santa Barbara–Los Angeles coast. We use the Navy’s Coupled Ocean/Atmosphere Mesoscale Prediction System^® (COAMPS) and COAMPS 4D-Variational methodology to integrate the HF radar-derived winds into the NWP initial conditions. Observation impact for the HF radar and all other observing platforms are quantified by forecast sensitivity observation impact (FSOI) and common model verification statistics generated by the Model Evaluation Toolkit.

The HF radar-derived wind observations have the potential to fill an observation gap near the shoreline where scatterometers from satellites are unable to observe and ship and buoy data are sparse. Work is already underway to test the HF Radar observations in other locations including the Chesapeake Bay.