Geophysical Model Function Research Articles

Influence of mesoscale eddies on wind retrieval from C-band synthetic aperture radar images

ABSTRACT Synthetic Aperture Radar (SAR) is an innovative technique for monitoring sea surface dynamics and has a large swath coverage (>200 km). In this paper, we investigate the characteristics of anticyclonic and cyclonic eddies on SAR and the influence of mesoscale eddies on SAR wind retrieval. Mesoscale eddies in the global oceans are identified from daily sea level anomaly (SLA) data in 2022–2023, which are integrated by the Haiyang-2B/2C/2D (HY-2B/2C/2D) altimeters. A total of 1000 Sentinel-1 (S-1) images acquired in interferometric-wide (IW) and extra-wide (EW) modes are available, and they cover the mesoscale eddies. In addition, swath wind products from HY-2 scatterometers and sea surface temperature (SST) data from Haiyang-1C (HY-1C) are collected. Co-polarized (vertical–vertical (VV) and horizontal–horizontal (HH)) Geophysical Model Function (GMF) C-SARMOD2 was applied to invert the wind speed with reference to the scatterometer-measured wind direction. The variation in the difference (SAR retrievals minus scatterometer measurements) indicates that the SST of below 11°C has a certain impact on the SAR wind retrieval. The influence is weak at higher SSTs, and the difference increases with increasing eddy kinetic energy (EKE). It is concluded that the shear current associated with mesoscale eddies is a significant factor that affects the accuracy of SAR-derived winds.

Sea surface wind retrieval from CFOSAT Rotating fan-beam scatterometer in ocean cyclones

ABSTRACT Rotating fan-beam scatterometer (RFSCAT) is a radar scatterometer system for sea surface wind vector measurement. Compared with other available scatterometers, RFSCAT can provide more combination of azimuth angles and incidence angles for a single WVC (wind vector cell), this observation mechanism is more conducive to the sea surface wind direction retrieval. In this paper, the NSCAT-4DS GMF (geophysical model function) with SST (sea surface temperature) correction, and the MSS (Multiple Solution Scheme) combination with a 2DVAR (2-dimensional variational analysis) are adopted to retrieve the sea surface winds from RFSCAT on the CFOSAT (Chinese-French Oceanography Satellite). The retrieved RFSCAT sea surface winds and ASCAT (advanced scatterometer) sea surface winds are compared and tested, and the feasibility of the RFSCAT measuring sea surface winds under high wind speeds is analysed. The results show that the RMSE of the RFSCAT sea surface wind speed using improved algorithm has decreased by 0.292 m s−1, the correlation coefficient has increased by 0.032, and the residual standard deviation has decreased by 0.194 m s−1. The RMSE of RFSCAT sea surface wind direction has decreased by 5.950°, the correlation coefficient has increased by 0.002, and the residual standard deviation has decreased by 5.567°. It is shown that the changes in winds RMSE using pre-improved and improved algorithms have statistical significance. In a word, the spaceborne Ku-band rotating fan-beam scatterometer can capture the winds structure of ocean cyclones. Although there may be high wind speeds saturation phenomena, the detecting sea surface winds at high wind speeds show preferable performance.

Deep residual fully connected network for GNSS-R wind speed retrieval and its interpretation

Global navigation satellite system reflectometry (GNSS-R) has emerged as a new technique to provide L-band bistatic measurements for ocean wind speed retrieval, in which traditional geophysical model functions (GMFs) or shallow neural networks (NNs) are normally used. However, it is still challenging to identify and consider all relevant parameters in the GMF. Meanwhile, NN models face limitations due to the degradation problem, which restricts their depth and consequently their performance. Furthermore, the interpretation of NN models for GNSS-R wind retrieval is another issue. To this end, we propose a residual fully connected network (RFCN) fusing auxiliary information such as geometry, receiver gain, significant wave height, and current speed with track-wise corrected σ0. Referred to the European Centre for Medium-Range Weather Forecast (ECMWF) ERA5 wind product, the root mean square error (RMSE) and bias of RFCN winds are 1.031 m/s and -0.0003 m/s, respectively, with a 6% improvement in RMSE compared to debiased NOAA Cyclone Global Navigation Satellite System (CYGNSS) Version 1.2 (V1.2) wind speed retrieval. Moreover, in an intertropical convergence zone (ITCZ) area with large current speeds, the RMSE and bias are 1.006 m/s and -0.022 m/s: an improvement of 11.6% and 87.9% compared to debiased NOAA CYGNSS V1.2 winds. The bias ‘strips’ in these areas are nearly eliminated. Daily averaged error analyses also demonstrate that RFCN winds are more robust and consistent with ECMWF winds. For wind speeds larger than 20 m/s, referred to Soil Moisture Active Passive (SMAP) Level 3 final wind products, the RMSE and bias of fine-tuning RFCN (FT_RFCN) winds are reduced by 25.7% and 91.5% compared to NOAA winds. Finally, the RMSE and bias of retrievals in tropical cyclones, measured by Stepped Frequency Microwave Radiometer (SFMR) during 2021-2022, reveal an improvement of 3.5% and 21.2% compared to NOAA winds. Through SHapley Additive exPlanations (SHAP) models developed for RFCN and FT_RFCN, the contribution of each feature is quantitatively evaluated, while providing insights into their interactions within the ‘black-box’ NN models with clear physical meanings.

Hybrid CMOD-Diffusion Algorithm Applied to Sentinel-1 for More Robust and Precise Wind Retrieval

Synthetic Aperture Radar (SAR) imagery presents significant advantages for observing ocean surface winds owing to its high spatial resolution and low sensitivity to extreme weather conditions. Nevertheless, signal noise poses a challenge, hindering precise wind retrieval from SAR imagery. Moreover, traditional geophysical model functions (GMFs) often falter, particularly in accurately estimating high wind speeds, notably during extreme weather phenomena like tropical cyclones (TCs). To address these limitations, this study proposes a novel hybrid model, CMOD-Diffusion, which integrates the strengths of GMFs with data-driven deep learning methods, thereby achieving enhanced accuracy and robustness in wind retrieval. Based on the coarse estimation of wind speed by the traditional GMF CMOD5.N, we introduce the recently developed data-driven method Denoising Diffusion Probabilistic Model (DDPM). It transforms an image from one domain to another domain by gradually adding Gaussian noise, thus achieving denoising and image synthesis. By introducing the DDPM, the noise from the observed normalized radar cross-section (NRCS) and the residual of the GMF methods can be largely compensated. Specifically, for wind speeds within the low-to-medium range, a DDPM is employed before proceeding to another CMOD iteration to recalibrate the observed NRCS. Conversely, a posterior-placed DDPM is applied after CMOD to reconstruct high-wind-speed regions or TC-affected areas, with the prior information from regions characterized by low wind speeds and recalibrated NRCS values. The efficacy of the proposed model is evaluated by using Sentinel-1 SAR imagery in vertical–vertical (VV) polarization, collocated with data from the European Centre for Medium-Range Weather Forecasts (ECMWF). Experimental results based on validation sets demonstrate significant improvements over CMOD5.N, particularly in low-to-medium wind speed regions, with the Structural Similarity Index (SSIM) increasing from 0.76 to 0.98 and the Root Mean Square Error (RMSE) decreasing from 1.98 to 0.63. Across the entire wind field, including regions with high wind speeds, the validation data obtained through the proposed method exhibit an RMSE of 2.39 m/s, with a correlation coefficient of 0.979.

Estimating effects of wind and waves on the Doppler centroid frequency shift for the SAR retrieval of ocean currents

Estimating the effects of wind and waves (EWW) on the Doppler centroid frequency shift is a key step in the Synthetic Aperture Radar (SAR) retrieval of ocean currents. However, available models for estimating EWW may exhibit errors as they fail to fully consider the wave and current effects or uncorrected electromagnetic pointing errors. In this study, we corrected the electromagnetic pointing error of Sentinel-1A. In particular, by matching the current data provided by the high-frequency (HF) radar, we accurately obtained the EWW from July 2020 to June 2023. By using the wind and wave parameters provided by the European Centre for Medium-Range Weather Forecasts (ECMWF), the variation characteristics of EWW were analyzed and four geophysical model functions (GMFs) were established using different wave parameters. The application of these GMFs to the retrieval of currents was validated using the HF radar. The optimal current retrieval accuracy in terms of the time series and statistical results was then achieved using the model containing the most comprehensive wave parameters, with a bias of 0.01 m/s and a root mean square error of 0.19 m/s. The proposed model demonstrates the ability to retrieve 100 km scale eddy currents in the Gulf Stream, exhibiting a high agreement with HYbrid Coordinate Ocean Model (HYCOM) reanalysis data, with a bias and root mean square error of approximately −0.03–0.13 m/s and 0.2–0.32 m/s, respectively. The results fully demonstrate that the GMFs established in this study can facilitate the improvement of the SAR retrievals of ocean currents. Moreover, the ability of the proposed model to capture the velocity variations of subscale eddy currents is proved, providing an important tool for oceanographic research.

Wind speed retrieval approach for VV-polarized synthetic aperture radar during tropical cyclone based on XGBoost model

ABSTRACT In this study, a machine learning approach is applied to wind speed retrieval from vertical-vertical (VV) polarized synthetic aperture radar (SAR) during tropical cyclones (TC). More than 2400 dual-polarized (VV and vertical-horizontal (VH)) Sentinel-1 (S-1) images acquired in interferometric wide (IW) and extra wide (EW) mode are collected in 2016–2022. Along-track observations from stepped-frequency microwave radiometer (SFMR) and soil moisture active passive (SMAP) are available for several images. The azimuthal cut-off wavelength (ACW) is derived from VV-polarized image and the Doppler centroid anomaly (DCA) is provided from Ocean (OCN) product. It is found that SFMR wind speed linearly relates with ACW and DCA. Following this rationale, a machine learning method, called eXtreme Gradient Boosting (XGBoost), is trained for constructing SAR wind speed retrieval geophysical model function (GMF) through abundant matchups from 2000 images collocated with wind speeds from soil moisture active passive (SMAP) microwave radiometer and SFMR, denoted as TCWIND2-S1. Then GMF TCWIND2-S1 is applied for more than 400 images and the validation against SMAP and SFMR products up to 70 m s−1 shows a 3.01 m s−1 root mean squared error (RMSE) with a 0.16 scatter index (SI) and a 0.92 correlation (COR). Retrieval is also compared with CyclObs wind products derived from more than 400 dual-polarized images, yielding a 2.66 m s−1 RMSE with a 0.19 SI and a 0.93 COR. Our work provides an approach for VV-polarized SAR wind retrieval at wind speed ranged from 0 to 70 m s−1 during TCs. This work is conveniently adopted for TerraSAR-X without operating at cross polarization channel.

Preliminary analysis of calibration for Chinese civilian 1mC-SAR

ABSTRACT The main purpose of this letter is to investigate the performance of the calibration of the Chinese civilian satellite carrying synthetic aperture radar (SAR), denoted as 1mC-SAR, which is a successor of the previous Gaofen-3 (GF-3) satellite. In total, more than 500 1mC-SAR images in co-polarization modes (vertical–vertical [VV] and horizontal–horizontal [HH]) are collocated with the wind products from Haiyang-2 (HY-2B/2C/2D) scatterometers in April – May 2023. The well-developed geophysical model function (GMF) CMOD7 and polarization ratio PR2011 in the C-band are used to simulate the normalized radar cross section (NRCS) employing the wind vector from HY-2. There are more than 2000 matchups for VV-polarization and more than 1000 matchups for HH-polarization. Comparison of the simulations and 1mC-SAR measurements yields a root mean squared error (RMSE) of 3.03 dB, a correlation coefficient (r) of 0.93, a scatter index (SI) of −0.14, and a bias of −2.22 dB for the VV-polarized NRCS, which is slightly better than the 3.68 dB RMSE, 0.90 r, −0.15 SI, and −3.08 dB bias of the HH-polarized NRCS. The variations in the difference in the VV-polarized NRCS (1mC-SAR measurements minus simulations) with respect to the wind speed and incidence angle are analysed. Despite the difference in the VV-polarization (i.e. <2 dB) at wind speeds of greater than 3 m s−1, the difference in the VV-polarization approaches 3 dB at several incidence angles, which needs to be further improved.

Tropical cyclone signatures in SAR ocean radial Doppler Velocity

Ocean surface radial Doppler Velocity (DV) signatures of Tropical Cyclones (TC) at moderate incidence angles are analyzed using a semi-empirical DV model. This model, originally named KaDOP, is based on the physics of Ka-band Doppler radar backscattering from the sea surface and represents the DV as a sum of components due to surface currents, long surface waves, Bragg scattering, and wave breaking. The latter two components were modified in this study to fit the model to C-band observations at strong winds. This modified model, referred to as a TC-DOP, requires several environmental input parameters, including TC winds and translation velocity, waves, and currents. For known winds and TC translation velocity, the surface currents are simulated using a model of the upper ocean response to TC passage. Waves are modeled using a self-similar description for the energy-containing TC-generated local wind waves and emanating swell. Given known winds, waves, and currents, the DV is calculated by the TC-DOP independently at each grid point. Calculations are performed for scanner-like observations that detect only the radial (cross-track) DV component. It is shown that at storm-strength winds and radar incidence angle, θ=45°, the main contribution to the total DV is provided by the surface current and wave breaking DV components (∼1 m s−1). The latter component has strong azimuth asymmetry and maximizes in the upwind sector. Next in magnitude is the contribution from wind waves (∼0.5 m s−1), followed by Bragg waves (∼0.2 m s−1), while the contribution from swell is less important. Comparisons of the TC-DOP calculations with the C-band Sentinel-1A SAR radial DV data show qualitatively good agreement. All observed radial velocity images in TC storm areas have a dipole-like feature similar to that predicted by the simulations, with a zero DV area roughly aligned with the cross-wind direction. The TC-DOP model prediction is found generally consistent with calculations utilizing an empirical geophysical model function (known as the CDOP). The residual differences are attributed to specific features of radar scattering mechanisms in different radar wavelength bands (Ka- vs C-band), as well as possible shortcomings of the Ka-band Modulation Transfer Function (MTF) originally derived from platform-based measurements.

Development of X-Band Geophysical Model Function for Sea Surface Wind Speed Retrieval with ASNARO-2

In the present study, a new geophysical model function (GMF) is developed for the X-band synthetic aperture radar (SAR) on board the Advanced Satellite with New System Architecture for Observation-2 (ASNARO-2) to retrieve accurate offshore wind speeds. Equivalent neutral wind speeds based on the local forecast model (LFM) are employed as reference wind vectors, and 12,259 matching points from 502 SAR images obtained with horizontal transmitting, horizontal receiving polarization around Japan are collected. To ensure convergence of the calculation, 8129 points are selected from the matching points to determine the basic formula for the GMF and 23 coefficients based on the relationships among the normalized radar cross section, wind speed, incidence angle, and relative wind direction. Compared with the reference wind speeds, the GMF wind speeds showed reproducibility with a bias of −0.10 m/s and an RMSD of 1.37 m/s. Additionally, it can be confirmed that the retrieved wind speed has the bias of 0.03 and the RMSD of 1.68 m/s when compared to the in situ wind speed from the Kuroshio Extension Observatory (KEO) buoy. The accuracy of these retrieved wind speeds is comparable to previous studies, and it is indicated that the developed GMF can be used to retrieve offshore winds from ASNARO-2 images.

A Technique for SAR Significant Wave Height Retrieval Using Azimuthal Cut-Off Wavelength Based on Machine Learning

A Technique for SAR Significant Wave Height Retrieval Using Azimuthal Cut-Off Wavelength Based on Machine Learning

Effects of ocean wave spectra on the polarized bidirectional reflectance distribution function matrix at Ku‐band and its implications on satellite backscattering simulations

A polarized bidirectional reflectance distribution function (pBRDF) matrix is developed from two-scale roughness theory with the aim of providing more accurate simulations of microwave emissions and scattering required for ocean–atmosphere coupled radiative transfer models. The potential of the pBRDF matrix is explored for simulating the ocean backscatter at Ku-band. The effects of ocean wave spectra including the modified Durden and Vesecky (DV2), Elfouhaily, and Kudryavtsev spectra on the pBRDF matrix backscatter simulations are investigated. Additionally, the differences in backscattering normalized radar cross-section (NRCS) simulations between the Ku-band geophysical model function and pBRDF matrix are analyzed. The results show that the pBRDF matrix can reasonably reproduce the spatial distribution of ocean surface backscattering energy, but the distribution pattern and numerical values are influenced by ocean wave spectra. The DV2 spectrum is the best one for the pBRDF matrix to simulate horizontally polarized NRCSs, with the exception of scenarios where the incidence angle is below 35°, the wind speed is less than 10 m s−1, and in the cross-wind direction. Also, the DV2 spectrum effectively characterizes the wind speed and relative azimuth angle dependence for vertically polarized NRCSs. The Elfouhaily spectrum is suitable for simulating vertically polarized NRCSs under conditions of low wind speed (below 5 m s−1) and incidence angles under 40°. The Kudryavtsev spectrum excels in simulating vertically polarized NRCSs at high incidence angles (> 40°) and horizontally polarized NRCSs at low incidence angles (< 35°).摘要为提供微波海洋-大气耦合辐射传输模型所需的精确发射和散射, 基于双尺度粗糙度理论发展了极化双向反射分布函数 (pBRDF) 矩阵. 本研究探讨了pBRDF矩阵是否能够表征ku波段洋面后向散射, 对比了三种常用海浪谱模型对pBRDF矩阵后向散射模拟的影响, 分析了ku波段地球物理模型函数 (GMF) 与pBRDF矩阵在后向散射归一化雷达截面 (NRCS) 模拟中的差异. 结果表明, pBRDF矩阵能够准确再现洋面后向散射能量的空间分布, 但后向散射能量分布受海浪谱影响较大.

Using Semicircular Sampling to Increase Sea Water/Ice Discrimination Altitude

The rapid development of aircraft and unmanned aerial vehicles (UAV) increases their use, including in polar areas, which are characterized by their remoteness and rather harsh conditions. The dominant trends in airborne radar development are expanding their functionality and increasing the altitude of their applicability. Our study focuses on the functionality enhancement of airborne high-altitude conical scanning radars currently used for circular clouds and precipitation observations as well as for sea wind measurements. Recently, we showed how a semicircular observation scheme, instead of a circular one, can double the maximum applicable altitude of sea wind measurements made with such radars. Here we apply this approach to show how an airborne high-altitude conical scanning radar’s functionality can also be expanded for sea water/ice discrimination within a semicircular observation scheme, again doubling the maximum discrimination altitude compared to circular observations. The discrimination is performed in scatterometer mode using the minimum statistical distance of the measured normalized radar cross sections (NRCSs) to the geophysical model functions (GMFs) of the sea water and ice underlying surfaces. However, as no sea ice GMF is available for the considered horizontal transmit and receive polarization at the Ku band, we instead used a substitute sea ice GMF having the same azimuth isotropic property setting for its NRCSs as the averaged value of the measured azimuth NRCSs within the semicircular observations scheme. Our analysis found that incidence angles of 30°, 45°, and 60° are well suited to our sea water/ice discrimination method, and that incidence angles higher than 30° are preferable as they provide a higher difference in the statistical distance of the measured NRCSs to the sea ice and water GMFs, whereas an incidence angle of 30° provides the highest applicable altitude for sea water/ice discrimination and wind retrieval. We also demonstrated the ability of the sea water/ice discrimination procedure’s implementation for any airborne wind scatterometer or multimode radar operated in scatterometer mode over freezing seas to avoid entirely erroneous sea wind measurement results when a sea ice surface is observed. The obtained results can also be used for enhancing aircraft and UAV radars and for developing new remote sensing systems. Doi: 10.28991/ESJ-2024-08-02-07 Full Text: PDF

ResNet Deep Learning‐Based Inversion Method for Sea Surface Wind Field

This paper presents a method to invert the sea surface wind field from Sentinel‐1 synthetic aperture radar (SAR) interferometric wide swath (IW) band mode ground range detected (GRD) vertical polarization (VV) data based on the deep residual network (ResNet). Compared with the traditional algorithm (i.e., using geophysical model functions (GMFs)), the wind field inversion based on the ResNet deep learning method solves the problem of separate wind direction and wind speed inversion. The ResNet model is developed based on the results of 212 VV polarized IW images and cross calibrated multiplatform (CCMP) data collected in 2021 in the offshore waters of China. The root mean square error, standard deviation, and bias of the inverse wind direction of the ResNet model are 12.5°, 9.8°, and 7.4°, respectively, and the root mean square error, standard deviation, and bias of the inverse wind speed are 2.33, 1.495, and −1.375. The ResNet is better than the conventional local gradient and C‐band model models, further confirming the advantages shown by the deep learning algorithm in nonlinear fitting and its potential application in sea surface wind field inversion.

Verification of C-Band Geophysical Model Function for Wind Speed Retrieval in the Open Ocean and Inland Water Conditions

Verification of the C-band geophysical model functions (GMF) for the open ocean and the Gorky reservoir was carried out using Sentinel-1 IW-mode Synthetic aperture radar (SAR) images. CMOD5.N, CMOD7, GMF for the Caspian Sea, and CSARMOD2 were considered. The motivation for this study is concerned with the clarification of applying C-band GMFs for SAR images including for the conditions of inland water bodies, as well as with the study of the influence of various wind speed direction sources on the results of wind speed magnitude retrieval for ocean conditions. Comparison of wind speed from the CMOD5.N algorithm using wind direction data from NOAA NDBC oceanographic buoys together with the data provided by NCEP reanalysis data showed that regardless of the geographic location, the result does not depend significantly on the choice of the wind direction source. Novel results of CMOD5.N, CMOD7, GMF for the Caspian Sea, and CSARMOD2 applications to the conditions of the Gorky reservoir are presented. The comparison of these results with the meteorological station measurements showed the best agreement for CMOD5.N. The preliminary results on the construction of new C-band GMF adjusted to the Gorky Reservoir have shown statistical parameters better than for Caspian Sea GMF and CSARMOD2.

Wind field reconstruction based on dual-polarized synthetic aperture radar during a tropical cyclone

ABSTRACT A wind field reconstruction method for dual-polarized (vertical-vertical [VV] and vertical-horizontal [VH]) Sentinel-1 (S-1) synthetic aperture radar (SAR) images collected during tropical cyclones (TCs) that does not require external information is proposed. Forty S-1 images acquired in interferometric-wide (IW) and extra-wide (EW) modes during the Satellite Hurricane Observation Campaign in 2015–2022 were collected. Stepped-frequency microwave radiometer (SFMR) observations made onboard the National Oceanic and Atmospheric Administration’s hurricane aircraft are available for 13 images. The geophysical model functions, namely VV-polarized C-SARMOD and cross-polarized S-1 IW/EW mode wind speed retrieval model after noise removal (S1IW.NR/S1EW.NR), were employed to invert the wind fields from the collected images. TC wind fields were reconstructed based on SAR-derived winds, enhancing TC intensity representation in the VV-polarized SAR retrievals and minimizing the error of the VH-polarized SAR retrievals at the sub-swath edge. The wind speeds retrieved from the SAR IW image were validated against the remote-sensing products from the soil moisture active passive (SMAP) radiometer, yielding a root mean squared error (RMSE) of approximately 4.3 m s−1, which is slightly smaller than the RMSE (4.4 m s−1) for the operational CyclObs wind product provided by the French Research Institute for Exploitation of the Sea (IFREMER). However, the CyclObs wind product has better performance than the approach proposed in this paper for the S-1 EW mode. Moreover, the RMSE of the wind speed between SAR-derived wind speed obtained using the proposed approach and the CyclObs wind product is within 3 m s−1 in all flow directions clockwise relative to north centered on the TC’s eye. This study provides an alternative method for TC wind retrieval from dual-polarized S-1 images without suffering saturation problem and external information; however, the pattern of the wind field around the TC’s eye needs to be further improved, especially at the head and back of the TC’s eye.

A COMPOSITE MODEL OF MICROWAVE SCATTERING FROM WATER SURFACE IN EXTREME WIND SPEED CONDITION

Experiments were carried out in the wind-wave flume of Large Thermo-Stratified Wind-Wave Tank of IAP RAS aimed at studying the mechanisms of cross-polarized microwave radiation scattering from water surface under conditions of extremely high wind speeds. It is shown that the normalized radar cross-section (NRCS) can be represented as the result of an incoherent addition of contributions from breaking wave crests and from non-breaking wind waves. The effect of smoothing the water surface after passing the breaking crest made it possible to measure the NRCS of the breaking area on cross-polarization, while no dependence of the NRCS on wind speed and incidence angle was revealed. NRCS on non-breaking wind waves was calculated within the framework of the small slope approximation (SSA) using experimentally measured wind wave spectra. It is shown that the NRCS on cross-polarization increases monotonically with increasing wind speed, including hurricane conditions. In this case, the contribution of non-breaking wind waves to the NRCS saturates at wind speeds above 25 m/s. The monotonous increasing NRCS at higher wind speeds is associated with a breaking area increasing. A composite model of microwave radiation scattering from wave-covered water surface has been constructed, which has been verified on the basis of comparison with measurement data. The possibility of constructing a geophysical model function for ocean conditions based on the proposed composite model is shown, which can be used for remote sensing of sea storms and hurricanes.

Novel Approach to Wind Retrieval from Sentinel-1 SAR in Tropical Cyclones

The strong winds in tropical cyclones (TCs) are commonly retrieved from cross-polarized SAR images using a geophysical model function (GMF). However, the accuracy of wind retrieval in cross-polarization is significantly reduced at the edges of sub-swaths. In this study, a novel approach to TC wind retrieval from VV polarized SAR images is proposed based on using the azimuthal cutoff wavelength to represent the effect of velocity bunching. A total of 12 dual-polarized (VV and VH) Sentinel-1 (S-1) images acquired in the interferometric wide (IW) mode were used, five of which were collocated with measurements taken by the Stepped-Frequency Microwave Radiometer (SFMR) on board an NOAA aircraft. The SAR-based azimuthal cutoff wavelengths were found to be linearly related to the SFMR wind speeds. Based on this finding, an empirical GMF for TC wind speed retrieval from VV S-1 images was constructed. The inversion results from seven images using this approach were validated against the wind products from the Advanced Scatterometer and the European Center for Medium-Range Weather Forecasts. The RMSE of the wind speed was 2.15 m s−1 and the correlation coefficient (COR) was 0.83 at wind speeds of less than 25 m s−1, while the RMSE was 2.66 m s−1 and the COR was 0.97 when compared with wind retrieval using the VH-polarized GMF S1IW.NR at wind speeds greater than 25 m s−1. The proposed algorithm performs well and has two advantages: (1) it is not subject to the saturation problem of the VV backscattering signal and (2) the discontinuity of the retrieval results obtained using VH GMF at the edges of sub-swaths is improved.

A New Approach for Ocean Surface Wind Speed Retrieval Using Sentinel-1 Dual-Polarized Imagery

A synthetic aperture radar (SAR) has the capability to observe ocean surface winds with a high spatial resolution, even under extreme conditions. The purpose of this work was to develop a new method for wind speed retrieval with the combination of SAR dual-polarized signals. In this study, we collected 28 tropical cyclone imageries observed using the Sentinel-1 dual-polarization mode. These imageries were collocated with radiometer wind speed measurements and reanalysis of wind vector products. In the new method, the wind speed was set as the output. VV-polarized (vertical transmitting–vertical receiving polarized) normalized radar cross section (NRCS), incident angle, VH-polarized (vertical transmitting–horizontal receiving polarized) NRCS, and wind direction were set as the inputs. Based on different output combinations, wind retrieval models were developed with multiple linear regression (MLR). According to the validation and comparison, the proposed models performed better than the traditional piecewise VH-polarization geophysical model functions (GMFs). The impact of thermal noise on the retrieval of low wind speeds (&lt;10 m/s) could be partially reduced. The input of wind direction is unnecessary if the combination of VV- and VH-polarized imageries has been utilized. These results suggest that the use of MLR and the dual-polarization combination can improve SAR wind retrieval accuracy. Compared with SMAP measurements, our SAR retrievals can provide fine structures of TC wind fields.

Machine Learning Applied to a Dual-Polarized Sentinel-1 Image for Wind Retrieval of Tropical Cyclones

In this work, three types of machine learning algorithms are applied for synthetic aperture radar (SAR) wind retrieval in tropical cyclones (TCs), and the optimal method is confirmed. In total, 30 Sentinel-1 (S-1) images in dual-polarization (vertical–vertical [VV] and vertical–horizontal [VH] were collected during the period from 2016 to 2021, which were acquired in interferometric-wide and extra-wide modes with pixels of 10 m and 40 m, respectively. More than 100,000 sub-scenes with a spatial coverage of 3 km are extracted from these images. The dependences of variables estimated from sub-scenes, i.e., VV-polarized and VH-polarized normalized radar cross-section (NRCS), as well as the azimuthal wave cutoff wavelength, on wind speeds from the stepped-frequency microwave radiometer (SFMR) and the soil moisture active passive (SMAP) radiometer are studied, showing the linear relations between wind speed and these three parameters; however, the saturation of VV-polarized NRCS and the azimuthal wave cutoff wavelength is observed. This is the foundation of selecting input variables in machine learning algorithms. Two-thirds of the collocated dataset (20 images) are used for training the process using three machine learning algorithms, i.e., eXtreme Gradient Boosting (XGBoost), Multi-layer Perceptron, and K-Nearest Neighbor, and the coefficients are fitted after training completion through 20 images collocated with SFMR and SMAP data. Another 10 images are taken for validation up to 70 m/s, yielding a 2.53 m/s root mean square error (RMSE) with a 0.96 correlation and 0.12 scatter index (SI) using XGBoost. The result is better than the &gt;5 m/s error achieved using the existing cross-polarized geophysical model function and the other two machine learning algorithms; moreover, the comparison between wind retrievals using XGBoost and Level-2 CyclObs products shows about 4 m/s RMSE and 0.18 SI. This suggests that the machine learning algorithm XGBoost is an effective method for inverting the TC wind field utilizing SAR measurements in dual-polarization.

ВОССТАНОВЛЕНИЕ СКОРОСТИ ПРИВОДНОГО ВЕТРА ПО ДАННЫМ МТВЗА-ГЯ

A neural network (NN) algorithm for the sea surface wind speed retrieval from the MTVZA-GYa Russian satellite microwave radiometer measurements is presented. The algorithm is based on the physical modeling of the brightness temperature of microwave radiation in the ocean-atmosphere system using new theoretical geophysical model functions of the dependence of ocean radiation on wind speed. The algorithm is validated by comparing the wind fields retrieved from the MTVZA-GYa data with those obtained from the AMSR2 radiometer (Japan) for different areas of the World Ocean with a difference in measurement time not exceeding five minutes. The validation has shown that the NNs with a number of neurons n from 3 to 8 provide the smallest root-mean-square difference between the AMSR2 and MTVZA-GYa retrieved wind speeds, namely 1.6 m/s. When mapping the wind speed in tropical cyclones, the best fit to the wind fields from the AMSR2 data is obtained using the NN with n = 4.