Use of satellite scatterometer winds to force an ocean general circulation model

Abstract

The present study is an attempt to evaluate the capacity of scatterometer wind data to be provided by the forthcoming European Remote Sensing 1 (ERS 1) and NASA Scatterometer (NSCAT) satellite missions, to successfully force basin scale ocean circulations. The mid‐latitude ocean is simulated with a three‐layer, quasi‐geostrophic, eddy‐resolving model in a square box geometry. The variable wind stress, which is used to simulate scatterometer winds, is derived from European Centre for Medium‐Range Weather Forecast (ECMWF) daily analyses of the wind stress over the North Atlantic. Its curl is applied as a forcing at every grid point of the model to drive a circulation of reference. The model is forced in real time under the swath with the curl derived from the raw scatterometer winds, with a condition of persistence that keeps the forcing constant between two consecutive passages of the satellite. The winds of ERS 1 (at 3‐ and 35‐day periods) produce circulations which globally resemble that of the reference experiment. A 3‐day repeat orbit appears to be the most suitable for producing raw winds capable of realistically driving a numerical ocean model. NSCAT (3‐day period) reproduces the main features of the reference large scale circulation with more success than ERS 1, and it is clear that NSCAT raw winds are far more suitable for driving an ocean Eddy‐Resolving General Circulation Model (EGCM) than ERS 1 winds because of their denser spatial coverage. ERS 1 and NSCAT raw wind stress curl fields are then objectively analyzed to produce smoothed “interpolated” wind stress curl fields having continuous values at every model grid point. Simulations with ERS 1 and NSCAT interpolated forcings show an improvement in the structure of the mean flow, but discrepancies in the energetics lead us to conclude that our interpolated winds are not significantly better than the raw winds. A «combined» forcing is constructed consisting of daily maps containing NSCAT raw data under the swaths and the corresponding objectively analyzed field outside the swaths. The circulation driven by the combined forcing reproduces remarkably well the main features and the energetics of the reference circulation. The patchy resolution of the combined forcing appeared to have little effect on the results of the simulations. Our conclusion is that, within the framework of the present study, scatterometer winds (and particularly those of NSCAT) seem appropriate for providing a forcing function to mid‐latitude ocean models, and for that purpose, these satellite data may not require prior assimilation in meteorological models, despite their patchiness and irregular resolution.

Sensitivity of the tropical Pacific Ocean simulation to the temporal and spatial resolution of wind forcing

“…The methodology of data processing is given by Liu et al [1998]. Here we are not concerned with the accuracy of the NSCAT data and the impact of sampling strategy, which have been the focus of many other studies [e.g., Barnier et al, 1991Barnier et al, , 1994Bourassa et al, 1997]. Suffice it to say that the NSCAT wind product is probably the best resolved wind data set available and its quality is sufficient for the purpose of this study.…”

“…They found that the mean and eddy kinetic energies could be enhanced by up to an order of magnitude in the eastern basin of the North Atlantic when the high-wavenumber forcing was applied. The sensitivity of ocean simulations to wind resolution has also been studied by investigators who were concerned with the impact of satellite scatterometer winds [Barnier et al, 1991[Barnier et al, , 1994.…”

The use of satellite scatterometer winds to drive a primitive equation model of the Indian Ocean: The impact of bandlike sampling

“…They conclude that more modeling studies are required before their results can be generalized. The present study extends the investigation of Barnier et al [ 1991]. Here a similar study is conducted but using a quite different model and a denser time sampling of the wind (6 hours instead of 1 day).…”

“…In a recent study using scatterometer wind data simulated from daily analyses of ECMWF, Barnier et al [1991] show that within the framework of quasi-geostrophic (QG), midlatitude ocean models, scatterometer winds from NSCAT (and to a lesser degree from ERS-1) would be appropriate to 14,187 Wave speed is 316 cm s -1 for baroclinic mode 1, 149 cm s -1 for baroclinic mode 2, and 100 cm s for baroclinic mode 3. force an ocean model. However, they stressed that their results are dependent upon the numerical model used (midlatitude, quasi-geostrophic) and upon the original wind field from which scatterometer winds were simulated (i.e., daily analyses from ECMWF).…”