Primary Title: HF-Radar Data Acquisition & Processing in the Gulf of Mexico
Three Project Titles
- University of Southern Mississippi – Gulf of Mexico Loop Current and Eddy Observations from High Frequency Radar Systems
- University of South Florida – Dry Tortugas and Lower Keys High Frequency Radars
- Texas A&M University – Passive Gulf of Mexico Loop Current Observations from High Frequency Radar Across the Yucatan Strait
Funding Agency: National Academies of Sciences, Engineering and Medicine (NASEM)
Project Lead: Scott Glenn
Partners: University of Southern Mississippi, University of South Florida, Texas A & M University, Codar Ocean Sensors, Shell
Period of Performance: 11/1/2018 – 12/31/2021
Total funding: $600,931
This project focuses on HF-Radar installation, operation, and data processing in the Gulf of Mexico, funded through three grants from NASEM. The systems will provide new, real-time data for model assimilation and validation to better understand the evolution of the Loop Current System.
This program will support HF-Radar (HFR) data flows along established paths within the IOOS network. Real-time data will flow through CORDC at Scripps into the U.S. National HF Radar DAC, an operational path operated by IOOS for over a decade. Post processing quality control of the HFR data and transfer to NCEI for archiving will be conducted at Rutgers by the protocols established by MARACOOS and approved by IOOS through the Regional Association certification process. The complete HFR dataset will be archived at GRIIDC.
USM will transfer the HFR data from the remote sites in the field to a central computer site at their University. Using standard and certified data flow pathways, the radial data files will be transferred from USM to CORDC hourly for real-time QARTOD quality control, production of total vector fields, and inclusion of the total vector files in the U.S. National HFR DAC and NDBC data webpage and THREDDS servers. Data also will be transferred between USM and Rutgers, a redundant hub on the National HFR backbone, and for post-processing quality control and submission to NCEI.
Post-processing is a human-in-the-loop process of retrospective lookbacks at the data and diagnostics to identify time periods that may require reprocessing of spectra with different SNR thresholds or different antenna patterns. This requires dedicated computers that can run the CODAR software, a computer server to store the original spectra, the time history of antenna pattern measurements, and the reprocessed radials and total vector products. It will include a THREDDS server to make post-processed radial and total vector files available to modelers as soon as possible. As already established by MARACOOS, a monthly transmission of reprocessed data to NCEI is envisioned.
The lookback process is more labor intensive for the Radars on the offshore platforms than for the land based radars because of the potential for changes in the antenna beam patterns. National IOOS standards recommend 1 pattern measurement per year, a number that may be sufficient for shore based sites with few moving metal objects nearby. Active weekly monitoring of site diagnostics is an IOOS-certified approach employed by MARACOOS to trigger additional pattern measurement responses independent of the annual revisits. But for the HF Radars on the offshore platforms, the baseline for frequency and magnitude of antenna pattern changes needs to be established site by site. Standard shore site pattern measurements by ship are expensive and by walking are impossible. For the offshore platforms we propose to run the CODAR AIS pattern measurement software weekly to establish a baseline for antenna pattern changes. The weekly changes in the patterns will be compared to changes in the systems diagnostics to establish triggers for the future.
The lookback process will also include weekly monitoring of the settings on the Signal to Noise Ratio (SNR) for which radial currents are flagged as suspect. Specifically, the global SNR ratio is used to define the first order Bragg peak that is input to the MUSIC algorithm. In the case of multi-modal current distribution, care must be taken to ensure this SNR includes all of the primary Bragg peak but excludes the 2nd order peaks used to process waves. The individual SNR produced by the CODAR radial metrics algorithms are then used to decide which of these radials to pass forward as good. The higher SNRs at this level have a dramatic influence on reducing the bearing uncertainty. All of the above procedures are currently used by MARACOOS for reprocessing radial currents before sending the highest quality data to NCEI.
As an additional post-processing QC validation check, the HFR current fields will be compared to available ADCPs, surface drifters and Sea Surface Temperature (SST) or ocean color satellite imagery. Most Quality Assurance of Real Time Oceanographic Data (QARTOD) recommendations treat every grid point as an independent time series, providing one layer of assessment to flag suspect data. Setting the QARTOD parameters to flag suspect HFR data will be aided by comparisons to ADCP and surface drifter data. Comparing maps of HFR data to SST or ocean color provides a second layer of assessment for the spatial maps. It is a valuable step at the inflow and outflow locations where multi-modal current distributions are expected and the Loop Current is a strong, persistent and identifiable feature. But more importantly, it is a critical step at the offshore platforms, when the oceanographic conditions are expected to switch from single modal current conditions when the Loop Current or Loop Current Eddy is not present, to multi-modal current conditions when the Loop Current or Loop Current Eddy is present. These map comparisons focused on the location of significant features enable identification of issues down to the range cell level, allowing corrective action to be taken including reprocessing to produce the highest quality data possible.