Real-Time Ocean Forecast System Verifications Using METplus

As part of its strategic plan toward a unified forecast system, NOAA/NWS/NCEP Environmental Modeling Center (EMC) is developing a real-time verification system to evaluate NCEP’s operational modeling suite using the METplus verification software package. The new EMC Verification System (EVS) will verify many of NCEP’s models operationally for the first time, including the Global Forecast System, Global Ensemble Forecast System, and Real-Time Ocean Forecast System (RTOFS), to provide model performance assessment in real time and identify potential model improvements through retrospective analysis. RTOFS is a high-resolution, eddy-resolving global hybrid coordinates ocean model with three-dimensional, multivariate data assimilation capability, which is initialized once a day and provides predictions of ocean currents, salinity, temperature and sea ice condition for up to eight days. The EVS RTOFS component generates real-time performance metrics and plots of mean error, root-mean-square error, and anomaly correlation coefficient for ocean temperature and salinity, sea surface height, and sea ice concentration. RTOFS forecasts are verified with multiple validation sources including satellite estimates, buoy data, and in-situ profiles of Argo floats. Performance metrics from the EVS are compared to those from a legacy RTOFS verification system running at EMC and show comparable results, albeit using different spatial interpolation methods and verification software. In this presentation, we will describe the EVS framework and configuration for the RTOFS component and highlight major results. We will also discuss the advantages of using METplus for real-time forecast verification, as well as technical challenges encountered with limited METplus capability to ingest ocean point observations and data on unstructured grids such as tripolar and polar stereographic grids.