An Overview of the Naval Research Laboratory Ocean Surface Flux (NFLUX) System

Jackie C May, Clark David Rowley and Charlie N. Barron, Naval Research Lab, Stennis Space Center, MS, United States
Abstract:
The Naval Research Laboratory (NRL) ocean surface flux (NFLUX) system is an end-to-end data processing and assimilation system used to provide near-real time satellite-based surface heat flux fields over the global ocean. Swath-level air temperature (TA), specific humidity (QA), and wind speed (WS) estimates are produced using multiple polynomial regression algorithms with inputs from satellite sensor data records from the Special Sensor Microwave Imager/Sounder, the Advanced Microwave Sounding Unit-A, the Advanced Technology Microwave Sounder, and the Advanced Microwave Scanning Radiometer-2 sensors. Swath-level WS estimates are also retrieved from satellite environmental data records from WindSat, the MetOp scatterometers, and the Oceansat scatterometer. Swath-level solar and longwave radiative flux estimates are produced utilizing the Rapid Radiative Transfer Model for Global Circulation Models (RRTMG). Primary inputs to the RRTMG include temperature and moisture profiles and cloud liquid and ice water paths from the Microwave Integrated Retrieval System. All swath-level satellite estimates undergo an automated quality control process and are then assimilated with atmospheric model forecasts to produce 3-hourly gridded analysis fields. The turbulent heat flux fields, latent and sensible heat flux, are determined from the Coupled Ocean-Atmosphere Response Experiment (COARE) 3.0 bulk algorithms using inputs of TA, QA, WS, and a sea surface temperature model field. Quality-controlled in situ observations over a one-year time period from May 2013 through April 2014 form the reference for validating ocean surface state parameter and heat flux fields. The NFLUX fields are evaluated alongside the Navy’s operational global atmospheric model, the Navy Global Environmental Model (NAVGEM). NFLUX is shown to have smaller biases and lower or similar root mean square errors compared to NAVGEM.