Abstract's details
Optimal SSH Mapping for Eddies and Mesoscale Currents
CoAuthors
Event: 2018 Ocean Surface Topography Science Team Meeting
Session: Others (poster only)
Presentation type: Type Poster
Contribution: not provided
Abstract:
Ocean Surface Current Analyses-Realtime currents (OSCAR, podaac.jpl.nasa.gov) are global ocean surface velocities calculated from sea surface height (SSH) gradients, ocean vector winds, and sea surface temperature fields using geostrophy, Ekman, and thermal wind dynamics. OSCAR uses the SSALTO/Duacs gridded MADT SSH fields (marine.copernicus.eu). OSCAR’s geostrophic currents have a strong correlation with drifting buoy velocities in highly geostrophic regions like the Gulf Stream, but even in these regions the amplitude is consistently lower than buoys. We have been using local polynomial fitting mapping methods to investigate the mesoscale and sub-mesoscale signal in altimetry and the effect that using gridded SSH fields has on the calculation of currents. An important feature of local polynomial fitting is the order of fit: a first-order (linear) fit provides the SSH gradients as part of the mapping.
The gridding method allows for a variable choice for the radius of points used for the fit. As the radius of points used in the gridding process is increased there is increased coverage for an individual map but the resulting velocities lose amplitude. Gapless coverage comes at the expense of accuracy and detail. What is interesting is that this gridding effect can explain the discrepancy in amplitude between OSCAR and drifter velocities only in some areas, and the structure of this dependence is regional. We explore the impact that the missing mesoscale geostrophic signals in OSCAR have on the circulation.
We also have been investigating these mapping methods on SSH fields. Optimal parameters for mapping T/P/J alongtrack data onto a regular grid are determined by sampling a high-resolution numerical model along the satellite tracks, then comparing the mapped fields with the original fields. This is done for a large number of different parameter settings, and shows that first-order fits perform systematically better than a zeroth-order fit, and variable bandwidth performs better than fixed bandwidth. These optimal parameters are then used to generate an open-source mapped altimeter product using only the T/P/J data.
Lastly, by accessing the high spatial resolution that is present in the along-track altimeter data we present the results of an along-track eddy census algorithm to study the global eddy field at scales as small as 20 km in some areas.
The gridding method allows for a variable choice for the radius of points used for the fit. As the radius of points used in the gridding process is increased there is increased coverage for an individual map but the resulting velocities lose amplitude. Gapless coverage comes at the expense of accuracy and detail. What is interesting is that this gridding effect can explain the discrepancy in amplitude between OSCAR and drifter velocities only in some areas, and the structure of this dependence is regional. We explore the impact that the missing mesoscale geostrophic signals in OSCAR have on the circulation.
We also have been investigating these mapping methods on SSH fields. Optimal parameters for mapping T/P/J alongtrack data onto a regular grid are determined by sampling a high-resolution numerical model along the satellite tracks, then comparing the mapped fields with the original fields. This is done for a large number of different parameter settings, and shows that first-order fits perform systematically better than a zeroth-order fit, and variable bandwidth performs better than fixed bandwidth. These optimal parameters are then used to generate an open-source mapped altimeter product using only the T/P/J data.
Lastly, by accessing the high spatial resolution that is present in the along-track altimeter data we present the results of an along-track eddy census algorithm to study the global eddy field at scales as small as 20 km in some areas.