This article deals with the calculation and validation of daily surface wind vector fields from wind speed and direction observations derived from Advanced SCATterometer (ASCAT) scatterometer measurements over the global ocean. According to the ASCAT sampling scheme, the objective method allowing for the determination of regular in space and time wind speed and direction fields uses ASCAT observations as well as European Centre for Medium Weather Forecasts (ECMWF) analyses. The latter are considered as external drift for the kriging method and as the temporal interpolation basis for ASCAT retrievals. This study focuses on the investigation of the capability of the method to add valuable wind information to the operational atmospheric analyses and on the quality of the resulting wind fields. The accuracy of the former is determined through comprehensive comparisons with daily winds calculated from moored buoy data. At global and regional scales, comparisons are performed with surface wind patterns derived from the ECMWF analysis and from ECMWF Re-analysis project (ERA-Interim) re-analyses.

A new method is developed to estimate daily turbulent air–sea fluxes over the global ocean on a 0.25° grid. The required surface wind speed (w ₁₀) and specific air humidity (q ₁₀) at 10 m height are both estimated from remotely sensed measurements. w ₁₀ is obtained from the SeaWinds scatterometer on board the QuikSCAT satellite. A new empirical model relating brightness temperatures (T _b) from the Special Sensor Microwave Imager (SSM/I) and q ₁₀ is developed. It is an extension of the author's previous q ₁₀ model. In addition to T _b, the empirical model includes sea surface temperature (SST) and air–sea temperature difference data. The calibration of the new empirical q ₁₀ model utilizes q ₁₀ from the latest version of the National Oceanography Centre air–sea interaction gridded data set (NOCS2.0). Compared with mooring data, the new satellite q ₁₀ exhibits better statistical results than previous estimates. For instance, the bias, the root mean square (RMS), and the correlation coefficient values estimated from comparisons between satellite and moorings in the northeast Atlantic and the Mediterranean Sea are –0.04 g kg⁻¹, 0.87 g kg⁻¹, and 0.95, respectively. The new satellite q ₁₀ is used in combination with the newly reprocessed QuikSCAT V3, the latest version of SST analyses provided by the National Climatic Data Center (NCDC), and 10 m air temperature estimated from the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalyses (ERA-Interim), to determine three daily gridded turbulent quantities at 0.25° spatial resolution: surface wind stress, latent heat flux (LHF), and sensible heat flux (SHF). Validation of the resulting fields is performed through a comprehensive comparison with daily, in situ values of LHF and SHF from buoys. In the northeast Atlantic basin, the satellite-derived daily LHF has bias, RMS, and correlation of 5 W m⁻², 27 W m⁻², and 0.89, respectively. For SHF, the statistical parameters are –2 W m⁻², 10 W m⁻², and 0.94, respectively. At global scale, the new satellite LHF and SHF are compared to NOCS2.0 daily estimates. Both daily fluxes exhibit similar spatial and seasonal variability. The main departures are found at latitudes south of 40° S, where satellite latent and sensible heat fluxes are generally larger.

Surface winds (10 m equivalent neutral wind velocity) from scatterometer missions since 1992 to present require homogenization to meet the requirements for oceanic and atmospheric climate data records. Sources of differences between winds retrieved from different scatterometer measurements mainly arise from calibration/validation procedures used for each scatterometer and differences in measurement physics. In this study, we focus on the calibration/validation component of the European Remote Sensing Satellite (ERS)‐1 and ERS‐2 wind speed biases. ERS‐1 and ERS‐2 data, named as WNF products, are from the Institut Français de Recherche pour l'Exploitation de la MER (IFREMER). In addition to WNF data, the newly calibrated ERS‐2 products provided by the European Space Agency (ESA), indicated as ASPS2.0 products, are also used. Our approach utilizes collocated satellite‐buoy data. Expected values of the normalized radar cross section (NRCS) are calculated from buoy winds for each antenna beam using the Cmod5.n geophysical model function. The comparisons between expected and measured NRCS examine differences along with variables such as backscatter coefficient and incidence angle ranges. The difference between the expected and measured NRCS is then used to set up empirical models aiming at the correction for biases in ERS‐1 and ERS‐2 WNF NRCS calibrations. Finally, ERS‐1 and ERS‐2 wind retrievals are reprocessed using the corrected NRCS and Cmod5.n. These earlier corrected ERS‐1/2 winds are analysed along with later scatterometer data (QuikSCAT and ASCAT‐A) for their deviations from in situ buoy winds during 1992–2011 period. The scatterometer data homogeneity is also investigated at global scales based on the use of collocated scatterometer retrievals and atmospheric re‐analyses winds derived from ERA Interim and CFSR models.

More than twelve satellite scatterometers have operated since 1992 through the present, providing the main source of surface wind vector observations over global oceans. In this study, these scatterometer winds are used in combination with radiometers and synthetic aperture radars (SAR) for the better determination and characterization of high spatial and temporal resolution of regional surface wind parameters, including wind speed and direction, wind stress components, wind stress curl, and divergence. In this paper, a 27-year-long (1992–2018) 6-h satellite wind analysis with a spatial resolution of 0.125° in latitude and longitude is calculated using spatial structure functions derived from high-resolution SAR data. The main objective is to improve regional winds over three major upwelling regions (the Canary, Benguela, and California regions) through the use of accurate and homogenized wind observations and region-specific spatial and temporal wind variation structure functions derived from buoy and SAR data. The long time series of satellite wind analysis over the California upwelling, where a significant number of moorings is available, are used for assessing the accuracy of the analysis. The latter is close to scatterometer wind retrieval accuracy. This assessment shows that the root mean square difference between collocated 6-h satellite wind analysis and buoys is lower than 1.50 and 1.80 m s⁻¹ for offshore and nearshore locations, respectively. The temporal correlation between buoy and satellite analysis winds exceeds 0.90. The analysis accuracy is lower for 1992–1999 when satellite winds were mostly retrieved from ERS-1 and/or ERS-2 scatterometers. To further assess the improvement brought by this new wind analysis, its data and data from three independent products (ERA5, CMEMS, and CCMP) are compared with purely scatterometer winds over the Canary and Benguela regions. Even though the four products are generally similar, the new satellite analysis shows significant improvements, particularly in the upwelling areas.

Scatterometer surface wind speed and direction observations in combination with radiometer wind speeds allow to generate surface wind analyses with high space and time resolutions over global as well as at regional scales. Regarding scatterometer sampling schemes and physics, the resulting surface wind analyses suffer from lack of accuracy in areas near coasts. The use of the synthetic aperture radar (SAR) onboard the Sentinel-1A satellite attempts to address the enhancement of surface wind analyses issues. In this study, SAR wind speeds and directions retrieved from backscatter coefficients acquired in interferometric wide (IW) swath mode are used. Their accuracy is determined through comprehensive comparisons with moored buoy wind measurements. SAR and buoy winds agree well at offshore and nearshore locations. The statistics characterizing the comparison of SAR and buoy wind speeds and directions are of the same order as those obtained from scatterometer (Advanced SCATterometer (ASCAT) and RapidScat) and buoy wind comparisons. The main discrepancy between SAR and buoy data are found for high wind speeds. SAR wind speeds exceeding 10 m s^–1 tend to be underestimated. A similar conclusion is drawn from SAR and scatterometer wind speed comparisons. It is based on the underestimation of SAR backscatter coefficient (σ°) with respect to σ° estimated from scatterometer winds and the geophysical model function (GMF) named CMOD-IFR2 (Ifremer C band MODel). New SAR wind speeds are retrieved using CMOD-IFR2. The corrected SAR retrievals allow better determination of the spatial characteristics of surface wind speeds and of the related wind components in near-coast areas. They are used for enhancing the determination of the spatial structure function required for the estimation of wind fields gridded in space and time at the regional scale. The resulting wind fields are only determined from scatterometer wind observations in combination with radiometer retrievals. Their qualities are determined through comparisons with SAR wind speeds and directions, and through their application for determination of wind power off Brittany coasts.