High resolution Sea Surface Wind retrieval over coastal Protected Areas by means of Sentinel-1 data

The algorithm used, i.e. SARWIND LG-Mod ver. v4.01 (see reference below), is aimed at producing the Sea Surface Wind (SSW), i.e. Speed and Direction, from a single co-polarized (VV or HH) SAR image. We used EW (Extended Wide) and IW (Interferometric Wide) Swath Mode GRD (Ground Range, Multi-Look, Detected) HR (High Resolution) Sentinel-1 images, with pixel spacings of 40m x 40m and 10m x 10m (azimuth x range) respectively. Associated auxiliary products were obtained from ESA SNAP 5.0 release. SSW fields were provided for the two coastal Protected Areas (PAs) named Camargue and Wadden Sea.

Each output folder of the SARWIND LG-Mod results contains useful plots and the estimated SSW field, provided in the file 'SAR_Sigma0_pp_decimationL2P2Tn_gradientOptSobel_LGMod_Results.txt' (pp = VV or HH; n = smoothing/decimation level), which is in the sub-folder 'LG-Mod_Theoretical_Results/Results_MEdegTHxx.xxx_Fisher (where xx.xxx is the final threshold applied). This txt file reports the following 19 columns:

1) LAT; 2) LON; 3) AZI; 4) RNG; [Location of the centre of the processed AOI]

5) REF_U; 6) REF_V; 7) REF_W; 8) REF_D; [ECMWF reference wind, as U/V components and speed/direction]

9) SAR_U; 10) SAR_V; 11) SAR_W; 12) SAR_D; [SARWIND LG-Mod wind estimates, as U/V components and speed/direction]

Both REF_D and SAR_D are wind directions (expressed in degrees) with respect to the geographic North (0°=North, 90°=East, 180°=South, 270°=West), that the wind is blowing to.
Both REF_W and SAR_W are wind speeds (expressed in m/s).
Regarding REF_U/SAR_U and REF_V/SAR_V, note that a positive U component represents wind blowing to the East; a positive V component represents wind blowing to the North.

13) SceneCentre_TrueHeading_FF; [Mean angle formed between the geographical South-North direction and the SAR azimuth direction (wrt the centre of the SAR Full-Frame image)]

SceneCentre_TrueHeading_FF is a positive clockwise angle. In particular: SceneCentre_TrueHeading_FF is in ]180,360[ [deg].
Thus:
Descending Pass <->  SceneCentre_TrueHeading_FF is in ]180,270[ [deg]
Ascending Pass  <->  SceneCentre_TrueHeading_FF is in ]270,360[ [deg]

14) ROI_Npoints_UnUsablePointsMasked; [Number of samples used for each SARWIND LG-Mod wind estimation]

15) MeanIncAng; 16) MeanNRCS; [Mean incident angle (expressed in degrees) and NRCS of the ROI]

17) MeanResultantLength; 18) Alpha2_Est; [Fisher's formula parameters]

19) MEdeg [Margin of Error, i.e. accuracy of each wind direction estimate, between 0° and 45°]

The accuracy MEdeg is given by the semi-width of the confidence interval, with a confidence level (1-α) fixed, which is assigned to the wind direction estimate. Consequently, lower MEdeg values correspond to better estimates. And, if MEdeg == 45°, wind estimates must be discharged.

Finally, note also that you can cut an entire row when [SAR_U SAR_V SAR_W SAR_D] == [NaN NaN NaN NaN] (typically, this happens for 'land pixels').

% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %
%                                                                                                                                       %
% REFERENCES:                                 %
%                                         %
% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %
%                                                                         %
% The algorithm SARWIND LG-Mod is based on the Ph.D thesis below:         %
%                                                                         %
% [1] Rana, Fabio Michele (2016) "Exploitation of Satellite               %
% Synthetic Aperture Radar Data for Geophysical Parameters Retrieval over %
% Land and Ocean". Unpublished Ph.D thesis. Politecnico di Bari.          %
%                                                                         %
% Some applications of the method are described in the following papers:  %
%                                                                         %
% [2] Fabio M. Rana, Maria Adamo, Guido Pasquariello, Giacomo De Carolis, %
% and Sandra Morelli, "LG-Mod: A Modified Local Gradient (LG) Method to   %
% Retrieve SAR Sea Surface Wind Directions in Marine Coastal Areas,"      %
% Journal of Sensors, vol. 2016, Article ID 9565208, 7 pages, 2016.       %
% doi:10.1155/2016/9565208.                                               %
%                                                                         %
% [3] Rana, F. M., Adamo, M., & Blanda, P. (2018, July).                  %
% LG-Mod Multi-Scale Approach for Sar Sea Surface Wind Directions         %
% Retrieval. In IGARSS 2018-2018 IEEE International Geoscience and Remote %
% Sensing Symposium (pp. 3216-3219). IEEE.                                %
%                                                                         %
% [4] Rana, F. M., Adamo, M., Lucas, R., & Blonda, P. (2019). Sea surface %
% wind retrieval in coastal areas by means of Sentinel-1 and numerical    %
% weather prediction model data. Remote Sensing of Environment, 225,      %
% 379-391.                                                                %
%                                                                         %
% Suggestions and comments are always welcome.                            %
% Thanks in advance,                                                      %
% Fabio Michele Rana                                                      %
%                                                                         %
% MOB: (+39) 3804114171                                                   %
% E-MAILS: fabiomichele.rana@gmail.com;  fabiomichele.rana@iia.cnr.it     %
%                                                                         %
% SKYPE: fabiomichelerana                                                 %
%                                                                         %
% SARWIND_LG-Mod_v4.01, 2014-2019                                         %
% Author: Fabio M. Rana                                                   %
% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %