Winds

    Roll vortices that are approximately aligned with the mean wind are a common feature of the turbulent atmospheric boundary layer (ABL) in near-neutral to moderately unstable stratification. This organized secondary circulation usually spans the depth of the ABL and forms bands of overturning circulations with alternating linear regions of enhanced (narrower and stronger) up-drafts and (weaker and broader) down-drafts forming between the counter-rotating roll circulations. Surface wind convergence is enhanced near the base of the updrafts and reduced at the bottom of the downdrafts. Similarly, the near-surface wind is reduced near the base of the updrafts and enhanced below the downdrafts. The net effect is an enhancement of the fluxes across the boundary layer that does not depend on the local mean vertical gradients. This non-local contribution to the turbulent fluxes is believed to play an important role in the overall air-sea exchange of momentum, heat and water vapor. Although the physics of rolls are becoming better understood through numerical and theoretical modeling, at present there is no global roll “climatology”. Parameterizing the contribution of rolls to the ABL fluxes requires numerous observations in order to characterize the structure of these rolls and the conditions under which they form and for which they do not. These are not easily obtained, especially over the world’s oceans.

    The Sentinel-1 constellation (S-1A & S-1B) launched by European Space Agency (ESA) in April of 2014 and 2016 routinely collects ∼120,000 (20 x 20 km, 5 m pixel) Synthetic Aperture Radar (SAR) wave mode vignettes in each month over open ocean. These SAR images provide routine and detailed sea surface imprints of many geophysical phenomena in nearly all weather conditions. When present, rolls usually induce a quasi-periodic pattern in the sea surface roughness that can be detected by SAR. The broad coverage of the S1 WV data provides a new and unique opportunity to characterize wind rolls at global scale.

    As a pilot study for a larger effort to characterize the atmospheric conditions associated with roll formation and consequent structure, a dataset consisting of 4800 S-1A SAR wave mode vignettes is manually created based on visual interpretation. These selected 4800 images with pure and clear patterns of wind rolls are uniformly distributed in 12 months of 2016. The horizontal wavelength and orientation of wind rolls of each vignette are extracted through the standard SAR image spectral analysis. Considering the vignette image size and the spectral characteristics of ocean swell, we restrict the horizontal wavelengths to the window 0.6-8 km. In addition, we collocated in time and space the SAR dataset with characteristic variables from ECMWF surface analyses. We seek the relationships between the near-surface stratifications (such as characterized by a bulk RIchardson number) and the roll characteristics. These results will be used to develop a strategy for determining an over-ocean roll climatology using routine S-1 observations with an eye toward improving PBL parameterization.

The possibility to get the wind field from the Synthetic Aperture Radar (SAR) images is extremely important in regional seas and coastal areas, only partially covered at present by satellite wind data. Here, the wind fields are often not well reproduced by both global and regional atmospheric models (Accadia and Zecchetto, 2007, Zecchetto and Accadia, 2014), probably because of the interaction between the wind flow and the orography and effects of the local sea water temperature. Furthermore, in small basins the atmospheric model winds are seldom validated, due to the lack of in-situ winds observations.

To retrieve the wind field from SAR two things are mandatory: a reliable radar-backscatter versus wind speed algorithm and a suitable estimate of the wind direction.

A method based on the 2D Continuous Wavelet Transform has been developed and applied to Sentinel-1 SAR images to retrieve the wind direction. The SAR derived wind directions have been compared with in-situ data the variability of which, expressed as the wind direction variance, has been found similar to that obtained from SAR (Zecchetto et al., 2016) and dissimilar to that from the WRF regional atmospheric model. Thus, it may be taken as a proxy for the reliability of the SAR wind directions.

The problem of the wind direction determination from SAR without any external information is thus virtually solved by the 2D-CWT method , and the resulting wind fields have been analysed and compared with those derived using WRF wind directions and with the OWI ESA products.

Possible uses of the SAR derived wind fields (wind vorticity and energy spectra) are also introduced and commented.

Unlimited access to L2 ocean wind fields from SAR opens up new possibilities for application of such wind products in connection with offshore wind energy exploitation. The European Space Agency (ESA) provides an Ocean Wind field component (OWI) retrieved from Sentinel-1 (S-1) observations. The Technical University of Denmark holds a comprehensive archive of SAR wind products for the European Seas generated in a systematic manner for both the Envisat and S-1 missions (see https://satwinds.windenergy.dtu.dk/). Institutions in the United States and Canada provide similar data offerings. End user’s access to wind maps and derived products tailored to wind energy applications has thus eased significantly.

This presentation shall focus on the value of a long-lasting collection of SAR observations for mapping of wind resources offshore. We first investigate the compatibility of wind fields retrieved from different European SAR sensors (Envisat, S-1A, S-1B) through comparisons with in situ observations of the wind speed. Any wind speed biases must be eliminated before the SAR wind data sets can be merged to a single time series.

Based on the merged SAR wind time series, statistical analyses are performed to estimate the mean wind speed, the Weibull distribution, and the energy density within grid cells of the dimension 0.02° latitude and longitude. This leads to detailed wind resource maps over the European seas including coastal waters. The advantage of SAR winds for offshore wind energy exploitation lies partially in the resolved coastal wind speed gradient. Gradients calculated from SAR winds compare well with ground based remote sensing wind measurements from the shoreline up to 3 km offshore. Wind speed gradients from SAR in coastal areas may be influenced by an increased uncertainty in the modelled wind direction, which we use as input for the SAR wind speed retrieval. Preliminary results suggest that small-scale local changes in the wind direction can introduce changing biases with the distance to shore.

Wind speeds retrieved from SAR are valid at the standard height 10 m above sea level. Wind resource maps for higher levels in the atmosphere, where wind turbines operate, are calculated on the basis of long-term average wind profiles given by a Numerical Weather Prediction (NWP) model. Wind resource maps based on SAR and modeling are made available as part of a larger mapping effort called the Global Wind Atlas (http://science.globalwindatlas.info/science.html) for the heights 10, 50, 100 and 150 m. The maps are updated as new S-1 SAR data is collected.

Altogether, synergies between SAR winds and outputs from NWP models are promising for wind resource mapping and wind farm planning in coastal areas where other types of observations are difficult and costly to gather. The open access to SAR wind maps and derived products represents an attractive new opportunity for wind energy developers and the use of SAR winds for planning of offshore wind farms could become standard practice in the near future.

This work is partially funded by the H2020 project CEASELESS (https://ceaseless.barcelonatech-upc.eu/en).

Abstract—A novel approach to ocean wind retrieval from high range bandwidth Synthetic Aperture Radar (SAR) data is demonstrated and validated using global Sentinel-1 (S1) a and b WV data acquired in October 2016 and January 2017. The first parameter, an integral spectral value (ISV), is defined from the high wavenumber area of the full resolution ocean image cross spectra. The second  paramter is the slope of the cross spectra phase plane (CSAPP). Together with the normalized radar cross section (NRCS) constitute the input to a data driven model for ocean wind speed and ocean wind direction retrieval system. The model is trained on S1B data and validated with S1A and S1B data co-located against ECMWF and Ascat data. The cross-spectral azimuth phase plane slope follows two sinusoidal functions, one symmetric and one anti-symmetric, with respect to the wind direction. The anti symmetric part is in direct relation to the azimuth wind direction, while the symmetric part indicates quadratic phase term related to surface acceleration effects. For wind speed standard deviation we achieve 1.54 m/s for S1A, with a bias of 0.18m/s.

Tropical cyclone is a generic term that designs a rapidly rotating storm system characterized by a low pressure center that produces strong winds and heavy rainfall. Cyclones are differently named according to the ocean basin where they develop: “Hurricanes” in the Northern Atlantic and Northeast Pacific Ocean, “Typhoons” in the Northwest Pacific Ocean, and “Cyclones” in the Indian Ocean and Southwest Pacific Ocean. Although tropical cyclones are among the most dangerous and destructive natural disasters, current models are still not able to give an accurate forecast of their intensity and track. Hence, within this context, spaceborne active microwave sensors, like the Synthetic Aperture Radar (SAR), are of paramount importance because of their all-weather and all-day capabilities together with a fine spatial resolution. In particular, SAR data can be used to directly monitor position and intensity of the storms, to analyze their structure or the rainfall-rate but also to estimate wind speed and direction. The purpose of this study is to apply the azimuth cut-off approach under extreme weather conditions.

In literature, the azimuth cut-off method is used to retrieve wind speed and several studies have been carried out to analyze the dependence of λc on sea surface parameters. In particular, there is a linear relationship between λc values and geophysical parameters, like wind speed under low-to-moderate regimes and significant wave height Recently, in [1] the ACF-based λc approach has been improved to deal with high wind speed regimes, e.g.; extreme weather conditions. The key issues that allow to extend the method to high wind regimes concern the tuning of the method with respect to pixel spacing, box size and the homogeneity of the SAR imagery. In particular, the box size is set at about 1 km × 1 km and an adaptive window size is selected for the median filter to account for the pixel spacing and first results on wind speed estimation under tropical cyclone conditions are presented.

In this study, an actual SAR dataset collected under tropical cyclones is used to discuss the soundness of this improved azimuth-cut-off method under extreme weather conditions .

 [1] M. Portabella, V. Corcione, X. Yang, Z. Jelenak, P. Chang, G. Grieco, A. Mouche, F. Nunziata, W. Li, “Analysis of the SAR-derived wind signatures over extra-tropical storm conditions”, Dragon 4 Symposium, Copenhagen, Denmark, 26-30 June.

Remote sensing techniques appear to be the most efficient way to monitor Tropical Cyclones (TC), as these low pressure systems evolve in data-sparse oceanic regions. In particular, C-Band SAR sensors are the only active microwave sensors able to observe ocean surface night and day and through clouds at high resolution (50 m) with a wide coverage (400 km swath). During summer 2016 (and in 2017), a campaign (SHOC) dedicated to the observation of hurricanes with Sentinel-1 was performed in close collaboration with ESA Mission Planning team in the frame of the ESA Scientific Exploitation of Operational Missions program.

First, this study presents the strategy adopted to adapt the acquisition planning to TC observations, and provide an overview of the hurricane database built with Sentinel-1 observations. Sentinel-1 is the first European SAR mission able to acquire data in wide swath modes with both co- and cross- polarizations. The quality and the potential of the signal acquired in cross-polarization are thus discussed. As observed, the signal in cross-polarization is found to more than 10 times sensitive than in co-polarization. Then, a new algorithm to combine co- and cross-polarized channels for ocean surface wind measurements over extremes is proposed and validated.

This study also shows how mandatory is the multi-sensors approach for such events. In particular, the benefit of having SMAP and SMOS radiometers for algorithm development is demonstrated. Co-located over extreme events, C-band co- and cross-polarized NRCS and L-band ocean surface roughness brightness temperature (TB,rough) are directly compared to analyze the similarities and differences between these two parameters at medium resolution (about 25 km). NRCS in VH and VV polarization (σ0,VH, σ0,VV) were acquired by Sentinel-1 C-band synthetic aperture radar. TB,rough is estimated from Brightness Temperatures (TB) measured by the L-band radiometer on-board Soil Moisture Active Passive (SMAP) mission. In situation of rain rate less than 20 mm/hr, a striking linear relationship is found between active C-Band cross-polarized NRCS and passive L-Band TB,rough. Compared to both high TB,rough and σ0,VH, co-polarized σ0,VV measurements saturate. As interpreted, this can correspond to a regime change of the air-sea interactions during extreme events.