Abstract's details

Observation-based estimates of ocean vertical covariances in the model unconstrained band

Joseph D'Addezio (Naval Research Laboratory, United States)


Gregg Jacobs (Naval Research Laboratory, United States)

Event: 2022 Ocean Surface Topography Science Team Meeting

Session: Science III: Mesoscale and sub-mesoscale oceanography

Presentation type: Type Oral

Recent increases in ocean observation density have led to innovations in ocean multi-scale data assimilation and modeling. One such dataset will be provided by the new wide-swath altimeter, SWOT. Fitting multiple horizontal scales has been the primary focus, generally leaving the associated scale-dependent vertical physics neglected. Here we present observation-based estimates of temperature and salinity depth-depth covariances in two different horizontal scales. Glider data from the recent S-MODE experiment were assimilated into an ocean simulation using a single-scale 3DVAR algorithm. We map the glider time series into distance covered allowing a Fourier analysis of model errors in wavenumber space, instead of frequency. Power spectral density of the model errors with respect to the gliders allow us to estimate a threshold horizontal scale for model skill. That value is found to be approximately 275 km. That value is used to develop a Gaussian kernel in order to partition the glider and model data into ‘constrained’ & ‘unconstrained’ distance-depth volumes. Depth-depth temperature and salinity covariances in each band are compared and contrasted. Glider and model covariances are found to be similar. The results have immediate utility as they can be used to: 1) provide an observation-based vertical covariance for the second step of a multi-scale assimilation and 2) derive and use model-based vertical covariances in regions where the unique datasets we have employed here are not immediately available.

Oral presentation show times:

Room Start Date End Date
Sala Grande Wed, Nov 02 2022,16:15 Wed, Nov 02 2022,16:26
Joseph D'Addezio
Naval Research Laboratory
United States