Conference Agenda

In this study, a novel algorithm to retrieve, from C-band synthetic aperture radar imagery, the current speed along the range direction under extreme sea state, is developed. To this aim, a Sentinel-1 dual-polarized synthetic aperture radar dataset consisting of 2300 images is collected during 200 tropical cyclones, whose eyes lie within 500 km from radar images. The dataset is complemented with collocated wave simulations from the Wavewatch-III model and reanalysis currents from the Hybrid coordinate ocean model. The corresponding tropical cyclone winds are officially released by the French Research Institute for Exploitation of the Oceans, while the Stokes drift following the wave propagation direction is estimated from the waves simulated using the Wavewatch-III model. In this study, first the dependence of wind, Stokes drift and range current on the Doppler centroid anomaly is investigated and, then, the extreme gradient boosting machine learning model is trained on 87% of the Sentinel-1 dataset for range current retrieval purposes. The remainder of the dataset is used for testing the retrieval algorithm, showing a root mean square error and a correlation coefficient of 0.11 m/s and 0.97, respectively, with the Hybrid coordinate ocean model outputs. A validation against measurements collected from two high frequency phased-array radars is also performed, resulting in a root mean square error and a correlation coefficient of 0.12 m/s and 0.75, respectively. Hence, experimental results confirm the soundness of the proposed range current speed inversion algorithm, although external wave data are needed.

Around 21:00 UTC on April 20th 2021, the Indonesian submarine (KRI Nanggala 402) lost contact while conducting a torpedo launch drill approximately 60 nautical miles north of the Bali Sea. This paper utilizes 256 satellite SAR (Synthetic Aperture Radar) datasets to analyze the spatiotemporal distribution characteristics of internal waves in the sea area near the Lombok Strait. Additionally, it involves the inversion of amplitude and phase velocity parameters for the internal wave stripes captured by Sentinel-1A/SAR two days earlier the submarine's disappearance, and the inversion results are verified using Himawari-8 satellite . Finally, based on the intensity of the tidal currents in the Lombok Strait, a rational analysis of the internal wave strength on the day of the submarine's sinking is conducted and the possibility of the internal wave causing the submarine sinking is discussed.

Lakes are regulators and sentinels of global climate change. Water clarity is a key physical lake variable, which can be used to track the long-term changes in lake ecosystem. The Qinghai-Tibetan Plateau (TP) is one of the most sensitive areas to global climate change, where lakes are rarely influenced by human activities. While the hydrological variations of Qinghai-Tibetan lakes were extensively studied, it has not yet well-understood how water optical variable and water quality changed in the Qinghai-Tibetan lakes under climate change. In this study, water clarity of Qinghai-Tibetan lakes was estimated from 1986 to 2021 based on the Landsat series data. Since the Landsat-5/TM imagery only has three bands in the visible domain, an updated QAA-RGB model was proposed to improve the applicability of the semi-analytical model on lakes with relatively low water clarity (< 1 m). To ensure the consistency of water clarity retrieval from Landsat series data, the MODIS data were also adopted to cross-calibrate the water clarity model. As a result, a significant increasing trend of TP lakes was revealed during the past 36 years (slope = 0.045m/yr，P<0.01) and larger growth was found in the western and northern TP. The relationship analysis between lake clarity and meteorological factors uncovered that the lake area expansion and NDVI raising were two dominant factors related to the lake clarity improvement on the TP, while the growth of temperature and precipitation may also contribute. In terms of sub-regions, lake area, NDVI and temperature were the most significant factor in driving water clarity rise during the past 36 years in western and northern TP, eastern TP, and southern TP respectively. This study indicated that the warming and humidification of the climate on TP had a significant effect on the lake clarity and lake ecosystem through lake volume and vegetation growth on the TP.

Harmful Algal Blooms (HABs) pose a great threat to human and animal health, their occurrence also has a significant impact on a variety of socio-economic and environmental factors. HAB events are now a global problem which effect food production, tourism, and ecosystem health.. It is expected that the occurrence of HABs is likely to grow significantly with the increase in human population coupled with climate change. Alexandrium minutum is a species of dinoflagellate which is known to be a producer of paralytic shellfish poisoning (PSP) and has a large-scale distribution impacting many aquatic systems around the world. It can manifest frequently at cell concentrations exceeding 10⁴cells/ L^-1 and has been observed to reach significantly higher concentrations, surpassing 10⁸ cells/ L^-1. This poses a considerable threat to coastal waters, especially in regions with intensive aquaculture activities.

This study will draw on satellite sensors which differ in spatial, spectral, and temporal resolution: Sentinel-2, Sentinel-3. The objectives of this study are to develop and validate HAB detection algorithms for near-shore coastal waters with better generalisation capability and lower computational overload that could improve the identification of the optical characteristics and spectral properties directly associated with Alexandrium minutum.

The research will be focused on four optically diverse regions of interest; The Danube Delta and Black Sea Coastline (Romania), Galician Coast (NW Spain), Shandong Peninsula Coast (China) and the Northern-South China Sea (China). Here, we will present results from the Galician coast and other European waters. We used in-situ biogeoptical data collected from dedicated field campaigns such as hyperspectral Remote Sensing Reflectance, Chlorophyll-a concentration, phytoplankton abundance and taxonomy along with particulate absorption to characterise the optical properties associated with blooms of Alexandrium minutum and developed an empirical algorithm to detect these blooms in near-shore coastal waters. An additional dataset from the weekly monitoring program conducted by the Technological Institute for the Control of the Marine Environment of Galicia (INTECMAR) was used to validate the algorithm using Sentinel-2 MSI and Senitnel-3 OLCI.

We tested 3 state-of-the-art atmospheric correction models (C2RCC, Polymer and Acolite) against in-situ hyperspectral data and evaluated their performance over coastal waters. We will present results on the optical characteristics of Alexandrium minutum and the potential of MSI and OLCI for their remote detection.

The SAR is an coherent imaging radar that allows for the acquisition of Earth observation using microwave frequencies, ensuring the ability to operate in all weather conditions.The SAR observations have contributed to a wide range of applications, both for military and civilian purposes, being of paramount importance for marine and maritime applications [1, 2].

All coherent imaging processes, such as those involved by SAR, are affected from speckle. It has been shown that once a tailored model is available [3,4], marine speckle, often associated with the form of multiplicative noise and indeed mitigated by use of multiple looking techniques, can be informative. From a physical viewpoint, the speckle (which is the low-pass filtering of the fading), arises from the fact that, for each rough resolution cell, the overall complex electromagnetic field Eˆ is the outcome of the coherent sum of Ns elementary complex contributes each characterized by a random amplitude and phase. As a result, the total received field can be mathematically modelled using a 2-D random walk, with independently and identically Gaussian distributed real and imaginary components; the amplitude has a Rayleigh probability distribution and the intensity has a negative exponential distribution[5]. The Rayleigh distribution model for the amplitude of backscattered signal, which is associated to the socalled fully developed speckle has been shown to fit well homogeneous land scenes and sea surface when a large area is illuminated by the radar or under a negligible long-wave modulation condition [6].

In this study, the Synthetic Aperture Radar (SAR) image speckle over marine scenes is modelled and tested over both C-band and X - band SAR imagery at variance of wind speed. A data set of Sentinel-1 single-look complex dual-polarimetric C-band and Cosmo Sky Med complex HH X-band SAR scenes spatially collocated, acquired under different wind regimes – from low to moderate - is processed using a twofold approach. On one side, the intensity of co-polarized speckle components, in X-band and C-band, is estimated against wind speed; on the other side, their statistical distributions are analysed by using normalized intensities moments (NIMs) are related to wind speed.

Promising results are obtained using showcases that refer to C-band and X-band SAR scenes collected under low-to-moderate wind regimes, with the aim of showing the different behavior of the speckle as a function of both wind speed and as the spatial resolutions of the sensors are presented.

References:

[1] C. R. Jackson and J. R. Apel, Eds., Synthetic Aperture Radar Marine Users Manual. Washington, DC: NOAA, 2004.

[2] F. Nunziata, X. Li, A. Marino, W. Shao, M. Portabella, X. Yang and A. Buono, ”Microwave satellite measurements for coastal area and extreme weather monitoring”, Remote Sensing, vol. 13, no. 16, pp. 3126, 2021.

[3] G. Ferrara, M. Migliaccio, F. Nunziata, and A. Sorrentino, “GK-based observation of metallic targets at sea in full-resolution SAR data: a multipolarization study,” IEEE Journal of Oceanic Engineering, vol.36, no.2, pp. 195-204, 2011.

[4] V. Corcione, A. Buono, F. Nunziata and M. Migliaccio, ”A sensitivity analysis on the spectral signatures of low backscattering sea Areas in Sentinel-1 SAR images”, Remote Sensing, vol. 13, pp. 1183, 2021.

[5] M. Migliaccio, G. Ferrara, A. Gambardella, F. Nunziata and A. Sorrentino, “A Physically Consistent Speckle Model for Marine SLC SAR Images,” IEEE Journal of Oceanic Engineering, vol. 32, n.4, pp. 839-847, 2007.

[6] P. Beckmann and A. Spizzichino, The Scattering of Electromagnetic Waves From Rough Surfaces. Norwood, MA: Artech House, 1963.

The C-band synthetic aperture radar (SAR) obser-vation is one of the most popular sources for high-resolution tropical cyclone (TC) wind speed estimation. The scarcity of high wind speed data with good quality restricts the inversion accuracy of high wind speed. It is a challenge to obtain a high-precision wind speed inversion model with sparse data points. In this article, the TC wind speed is successfully estimated from the dual-polarization SAR signal. First, the training dataset with a total of 327 data points is formed using the Sentinel-1A (S-1A) extra-wide / interferometric-wide mode images and their temporal–spatial matched stepped frequency microwave radiometer (SFMR) measurements. Then, a novel Nonparametric TC Wind Speed Estimation model (hereafter NWSE model) is proposed with this dataset by using the Bayesian nonparametric general regression method. The wind speed is interpreted as a function of the cross-polarized normalized radar cross sections and incident angle for the NWSE model. Moreover, the wind speed obtained from the co-polarized signal is used to improve the accuracy of NWSE model under low wind speed. Finally, the validation results show the excellent overall consistency between the model retrieved wind speed and the collocated SFMR and Soil Moisture Active and Passive (SMAP) measurements. Specifically, considering all the wind speeds, the overall root-mean-square error (RMSE) and the absolute bias of NWSE model are 2.85 and 2.26 m/s compared with the SMAP wind speed, respectively. When considering the wind speeds larger than 30 m/s, the RMSE and bias of NWSE model are 3.75 and 2.78 m/s, respectively.

Coastal macroalgae blooms profoundly influence coastal marine ecology, tourism, and aquaculture. Since 2008, the western coasts of the Yellow Sea have been damaged by green tide caused by the overgrowth of Ulva prolifera every summer.

Remotely sensed optical radiation has been exploited to detect algae and to track them over time [1,2]. Recently, the all-day and almost all-weather observation ability provided by measurements remotely sensed by synthetic aperture radar (SAR) has been investigated to monitor the green tide. Mostly, phenological studies have been carried out that look at the gray-tone SAR image without any physical characterization of the mechanisms that generate unique algae signature in the SAR image plane [3].

In this study, an electromagnetic perspective on the unique algae signatures observed in co- and cross-polarized C-band SAR imagery collected by the Copernicus Sentine-1 platform is provided. The analysis distinguishes four showcases that call for peculiar co- and cross-polarized signatures and investigates the role of surface and volume components in the backscatter signal and their strength with respect to the Sentinel-1 noise equivalent sigma zero (NESZ) that are mostly collected under calm sea conditions, i.e., a challenging scenario in terms of NESZ.

The processing chain consists of using a deep learning (DL) model GA-Net to detect algae [3]. The detected algae maps are investigated in terms of the surface and volumetric scattering component they are responsible for and a new metric that provides info comparable to the conventional degree of polarization without using the phase information (i.e., suitable to work on GRD amplitude-only SAR data) is proposed.

Experimental results show that at C-band algae are mostly dominated by surface scattering with a residual cross-pol component that may be related to either misaligned multiple reflection or volumetric scattering.

References:

[1] Hu, C., Qi, L., Hu, L., Cui, T., Xing, Q., He, M., ... & Wang, M. (2023). Mapping Ulva prolifera green tides from space: A revisit on algorithm design and data products. International Journal of Applied Earth Observation and Geoinformation,116, 103173.

[2] Gao, L., Li, X., Kong, F., Yu, R., Guo, Y., & Ren, Y. (2022). AlgaeNet: A deep-learning framework to detect floating green algae from optical and SAR imagery. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15, 2782-2796.

[3] Guo, Y., Gao, L., & Li, X. (2022). A deep learning model for green algae detection on SAR images. IEEE Transactions on Geoscience and Remote Sensing, 60, 1-14.

We developed a scalable deep-learning model, the AlgaeNet model, for floating Ulva prolifera (U. prolifera) detection in moderate-resolution imaging spectroradiometer (MODIS) and synthetic aperture radar (SAR) images. We labeled 1055/4071 pairs of samples, among which 70%/30% were used for training/validation. As a result, the model reached an accuracy of 97.03%/99.83% and a mean intersection over a union of 48.57%/88.43% for the MODIS/SAR images. The model was designed based on the classic U-Net model with two tailored modifications. First, the physics information input was a multichannel multisource remote sensing data. Second, a new loss function was developed to resolve the class-unbalanced samples (algae and seawater) and improve model performance. In addition, this model is expandable to process images from optical sensors (e.g., MODIS/GOCI/Landsat) and SAR (e.g., Sentinel-1/GF-3/Radarsat-1 or 2, reducing potential biases due to the selection of extraction thresholds during the traditional threshold-based segmentation. We process satellite images containing U. prolifera in the Yellow Sea and draw two conclusions. First, adding the 10-m high-resolution SAR imagery shows a 63.66% increase in algae detection based on the 250-m resolution MODIS image alone. Second, we define a floating and submerged ratio number (FS ratio) based on the floating and submerged parts of U. prolifera detected by SAR and MODIS. A research vessel measurement confirms the FS ratio to be a good indicator for representing different life phases of U.prolifera

The Synthetic Aperture Radar (SAR), owing to its day-night and almost all-weather imaging capabilities together with its fine spatial resolution, is a valuable tool to observe the oceans and monitoring oil pollutions. Mineral oil films appear in the SAR image plane as spots darker than the sea surface background because of their suppression of gravity-capillary waves. This study contrasts predicted X-band sea surface backscattering from slick-free and oil-covered sea surfaces with actual measurements acquired by the X-band satellite TerraSAR-X (TSX) and COSMO-SkyMed (CSK) SAR missions. Two SAR scenes were acquired with a temporal difference of about 36 minutes, under similar met-ocean conditions, during the Gannet Alpha oil spill accident occurred in the North Sea. The normalized radar cross section of the slick-free sea surface is predicted using the Advanced Integral Equation Model (AIEM) while the backscatter from the oiled sea surface is predicted by the AIEM augmented with the Model of Local Balance (MLB) to include the damping effect of oil slicks.

Experimental results show that X-band co-polarized numerical predictions agree reasonably well with both TSX and CSK actual measurements collected over slick-free sea surface. When dealing with oil-covered sea surface, the predicted backscattering shows a fairly good agreement with TSX measurements, while it overestimates the CSK ones. This is likely due to the different spreading conditions of the oil imaged by the two satellite missions.

This study is supported by the ESA-NRSCC Dragon-5 cooperation project “Monitoring harsh coastal environments and ocean surveillance using radar remote sensing sensors” (ID 57979).

The 1-D ocean wave spectra provide good information for the roughness in remote sensing observations and simulations. The Surface Waves Investigation and Monitoring instrument (SWIM）onboard the China France Oceanography Satellite (CFOSAT) generates wave descriptions in the wave-number range of 0.01~0.25 rad/m (wavelength range: 30-500 m). Though mainly for long swell, in the widely existed developing sea-state, wind casts also affects in the roughness, while up-to-date spectra generally based on the fully developed condition. In this research, the differences in the generally applied spectra, i.e. the Goda and Elfouhaily spectrum are analyzed and compared with the SWIM ocean spectrum measurements. Results show that the Goda specptrum fits the observations better than the Elfouhaily spectrum, due to wave-number settings. However, to fit the developing sea state description with contributions from both wind and swell, we integrated in the amplitude dimension in the Goda spectrum expressions the terms from Elfouhaily spectrum, and derived parameters from the SWIM products. Specifically, wind speed and inverse wave age as well as the spectral peak features of the SWIM measured products are applied. When compared with the SWIM products not included in the parameter derivation, the combined spectrum shows better fitting with the SWIM 1-D spectrum product than Goda spectrum alone in the metrics from the difference index (DI) and R-Square (R²). Then the real sea states vary with wind speed, dominant wavelength, inverse wave age and wave steepness better the two existing spectrum models in the amplitude dimension have been obtained for descriptions from SWIM that is closer to the nature environment. Further research would be work on the improvement of the directional description.

In this study, the discernible disparity between cross-polarized (HV/VH) and co-polarized (HH/VV) normalized radar backscatter cross-section (NRCS) from the sea surface is observed, particularly under wind speeds exceeding 20 m/s. The absence of saturation in the cross-polarized NRCS, as different from the observed phenomenon in the co-polarized NRCS, makes cross-polarized synthetic aperture radar (SAR) imagery a favorable choice for high wind speed observation. Additionally, this investigation presents a novel model designed to describe the complex relationships among C-band cross-polarized NRCS, wind velocity, and the radar's angle of incidence. Utilizing a collection of sixteen ScanSAR wide mode images captured by RADARSAT-2 (RS-2) during tropical cyclone events, complemented by corresponding wind speed data from the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Stepped-Frequency Microwave Radiometer (SFMR). As shown by SAR data, the cross-polarized NRCS monotonically increases with wind speeds. At moderate winds, the cross-polarized NRCS is most sensitive to wind speed. However, at lower and higher wind speeds, the values of the cross-polarized NRCS increases slowly with wind speed. Meanwhile, the analysis also shows that the relationship between NRCS and wind direction is not obvious, but it has a tendency to decrease with the increase of incidence angle. Finally, the RS-2 data are divided into four groups according to the sub-swath of the ScanSA and the wind speed retrieval models are fitted respectively. The retrieval results are compared with the ECMWF wind speed below 22 m/s. The root mean square error (RMSE) is 1.89 m/s and the bias is 1.45 m/s. Compared with the wind speed measured by SFMR above 22 m/s, the RMSE and bias are 3.09 m/s and 2.62 m/s, respectively. Compared with spaceborne L-band radiometer onboard Soil Moisture Active Passive (SMAP), this model still reaches a good accuracy. When applying the wind speed retrieval model to 32 extra-wide-swath (EW) mode images from Sentinel-1A/B under tropical cyclone conditions, the RMSE and bias are 3.67 m/s and 2.77 m/s, respectively, compared with SMAP radiometer measurements. This demonstrates further the model's robustness and reliability in high wind speed estimation.