|Paper title||Retrieving QuikSCAT winds closer to the coast|
|Form of presentation||Poster|
High resolution accurate coastal winds are of paramount importance for a variety of applications, both civil and scientific. For example, they are important for monitoring some coastal phenomena such as orographic winds, coastal currents and the dispersion of atmospheric pollutants, or for the deployment of off-shore wind farms. In addition, they are fundamental for improving the forcing of regional ocean models and, consequently, the forecasting of some extreme events such as the Acqua Alta often occurring in the Venice lagoon.
Scatterometer-derived winds represent the golden standard. However, their use in coastal areas is limited by the land contamination of the backscatter Normalized Radar Cross Section (NRCS) measurements. Nonetheless, the coastal sampling may be improved if the Spatial Response Function (SRF) orientation and the land contamination are properly considered in the wind retrieval processing chain.
This study focuses on improving the coastal processing of the Seawinds scatterometer onboard QuikSCAT as part of a EUMETSAT study in the framework of the Ocean Sea Ice Satellite Application Facilities (OSI-SAF).
In particular, the analytical model of the SRF is implemented with the aim of computing the so-called Land Contribution Ratio (LCR), which is, by definition, the portion of the footprint area covered by land. This index is then used for a double purpose: a) removing the excessively contaminated measurements; b) implementing a LCR-based NRCS correction scheme for the relatively low contaminated measurements. A second SRF estimate is obtained from a pre-computed Look-Up Table (LUT) of SRFs that are parameterized with respect to (w.r.t.) the orbit time, the latitude of the measurement centroid and the azimuth antenna angle.
Finally, the useful measurements (including those LCR-based corrected) are averaged in order to obtain integrated measurements by beam or view, which are then input in the wind field retrieval processor. Two different averaging procedures, i.e., a box car and a noise-weighted averaging, are implemented
A detailed comparison between the anaytical and the LUT-based SRF models is shown and the consistency of the derived LCR indices is verified against the coastline. A sensitivity analysis of the LCR-based NRCS correction scheme w.r.t. the LCR threshold is carried out. The effects of both averaging procedures on the retrieved winds are carefully analyzed. Finally, the retrieved winds are validated against some coastal buoys and their accuracy is assessed. Preliminary results will be presented and discussed at the conference.