Search

Haze or cloud always shrouds satellite images, obscuring valuable geographic information for military surveillance, natural calamity surveillance and mineral resource exploration. Satellite image dehazing (SID) provides the possibility for better applications of satellite images. Most of the existing dehazing methods are tailored for natural images and are not very effective for satellite images with non-homogeneous haze since the semantic structure information and inconsistent attenuation are not fully considered. To tackle this problem, this study proposes a hierarchical semantic-guided contextual structure-aware network (SCSNet) for spectral satellite image dehazing. Specifically, a hybrid CNN–Transformer architecture integrated with a hierarchical semantic guidance (HSG) module is presented to learn semantic structure information by synergetically complementing local representation from non-local features. Furthermore, a cross-layer fusion (CLF) module is specially designed to replace the traditional skip connection during the feature decoding stage so as to reinforce the attention to the spatial regions and feature channels with more serious attenuation. The results on the SateHaze1k, RS-Haze, and RSID datasets demonstrated that the proposed SCSNet can achieve effective dehazing and outperforms existing state-of-the-art methods. [ABSTRACT FROM AUTHOR]

As an important food crop, summer maize is widely planted all over the world. Monitoring its growth and output is of great significance for world food security. With the trend of global warming and deterioration, the frequency of high temperature and heat damage affecting summer corn has been increasing in the past ten years. Therefore, there is an increasing demand for monitoring the high temperature and heat damage of summer maize. At present, there are nearly a hundred indices or methods for research on high temperature and heat damage. However, research based on the vegetation index cannot fully describe the damage caused by high-temperature thermal damage, and there is an obvious asynchrony effect. Research based on hyperspectral remote sensing has many inconveniences in data acquisition and complex physical model construction. Therefore, this study uses remote sensing data, including MODIS surface reflection data, MODIS land surface temperature products, as well as ground observation data and statistical data, combined with multiple remote sensing indices and land surface temperature, to construct a remote sensing index, LSHDI (land surface heat damage index). The LSHDI first searches for a location with the worst vegetation growth conditions in the three-dimensional feature space based on the LST (land surface temperature), the normalized difference vegetation index (NDVI), and the land surface water index (LSWI). Then, it calculates the distance between each point and this location to measure the degree of vegetation affected by high temperature and heat damage. Finally, because there is no reliable disaster verification dataset that has been published at present, this study uses soil moisture as a reference to explain the performance and stability of the LSHDI. The results showed that their coefficient of determination was above 0.5 and reached a significance level of 0.01. The LSHDI can well-reflect the high temperature and heat damage of land surface vegetation and can provide important data support and references for agricultural management departments. [ABSTRACT FROM AUTHOR]

Global navigation satellite system reflectometry (GNSS-R) is a remote sensing technology of soil moisture measurement using signals of opportunity from GNSS, which has the advantages of low cost, all-weather detection, and multi-platform application. An in situ GNSS-R and radiometer fusion soil moisture retrieval model based on LSTM (long–short term memory) is proposed to improve accuracy and robustness as to the impacts of vegetation cover and soil surface roughness. The Oceanpal GNSS-R data obtained from the experimental campaign at the Valencia Anchor Station are used as the main input data, and the TB (brightness temperature) and TR (soil roughness and vegetation integrated attenuation coefficient) outputs of the ELBARA-II radiometer are used as auxiliary input data, while field measurements with a Delta-T ML2x ThetaProbe soil moisture sensor were used for reference and validation. The results show that the LSTM model can be used to retrieve soil moisture, and that it performs better in the data fusion scenario with GNSS-R and radiometer. The STD of the multi-satellite fusion model is 0.013. Among the single-satellite models, PRN13, 20, and 32 gave the best retrieval results with STD = 0.011, 0.012, and 0.007, respectively. [ABSTRACT FROM AUTHOR]

Ballistic target recognition is of great significance for space attack and defense. The micro-motion features, which contain spatial and motion information, can be regarded as the foundation of the recognition of ballistic targets. To take full advantage of the micro-motion information of ballistic targets, this paper proposes a method based on feature fusion to recognize ballistic targets. The proposed method takes two types of data as input: the time–range (TR) map and the time–frequency (TF) spectrum. An improved feature extraction module based on 1D convolution and time self-attention is applied first to extract the multi-level features at each time instant and the global temporal information. Then, to efficiently fuse the features extracted from the TR map and TF spectrum, deep generalized canonical correlation analysis with center loss (DGCCA-CL) is proposed to transform the extracted features into a hidden space. The proposed DGCCA-CL possesses better performance in two aspects: small intra-class distance and compact representation, which is crucial to the fusion of multi-modality data. At last, the attention mechanism-based classifier which can adaptively focus on the important features is employed to give the target types. Experiment results show that the proposed method outperforms other network-based recognition methods. [ABSTRACT FROM AUTHOR]

Scientometric reviews, facilitated by computational and visual analytical approaches, allow researchers to gain a thorough understanding of research trends and areas of concentration from a large number of publications. With the fast development of satellite altimetry, which has been effectively applied to a wide range of research topics, it is timely to summarize the scientific achievements of the previous 50 years and identify future trends in this field. A comprehensive overview of satellite altimetry was presented using a total of 8541 publications from the Web of Science Core Collection covering the years from 1970 to 2021. We begin by presenting the fundamental statistical results of the publications, such as the annual number of papers, study categories, countries/regions, afflictions, journals, authors, and keywords, in order to provide a comprehensive picture of satellite altimetry research. We discuss the co-occurrence of the authors in order to reveal the global collaboration network of satellite altimetry research. Finally, we utilised co-citation networks to detect the development trend and associated crucial publications for various specific topics. The findings show that satellite altimetry research has been changed immensely during the last half-century. The United States, France, China, England, and Germany made the most significant contributions in the field of satellite altimetry. The analysis reveals a clear link between technology advancements and the trend in satellite altimetry research. As a result, wide swath altimetry, GNSS-reflectometry, laser altimetry, terrestrial hydrology, and deep learning are among the most frontier study subjects. The findings of this work could guide a thorough understanding of satellite altimetry's overall development and research front. [ABSTRACT FROM AUTHOR]

Biomass burning (BB) is a driving force for heavy haze in northeastern China (NEC) and shows distinct seasonal features. However, little is known about its climatological aspects, which are important for regional BB management and understanding BB effects on climate and environment. Here, the climatological characteristics of active fires and their dependence on land cover in NEC were studied using Moderate Resolution Imaging Spectroradiometer (MODIS) products. Moreover, the influence of meteorological factors on fire activities was explored. The number of fires was found to have increased significantly from 2003 to 2018; and the annual total FRP (FRPtot) showed a generally consistent variation with fire counts. However, the mean fire radiative power for each spot (FRPmean) decreased. Fire activity showed distinctive seasonal variations. Most fires and intense burning events occurred in spring and autumn. Spatially, fires were mainly concentrated in cropland areas in plains, where the frequency of fires increased significantly, especially in spring and autumn. The annual percentage of agricultural fires increased from 34% in 2003 to over 60% after 2008 and the FRPtot of croplands increased from 12% to over 55%. Fires in forests, savannas, and grasslands tended to be associated with higher FRPmean than those in croplands. Analysis indicated that the increasing fire count in NEC is mainly caused by agricultural fires. Although the decreasing FRPmean represents an effective management of BB in recent years, high fire counts and FRPtot in croplands indicate that the crop residue burning cannot be simply banned and a need instead for effective applications. More efforts should be made on clean utilization of straw. The accumulation of dry biomass, high temperature, and low humidity, and weak precipitation are conducive to the fire activities. This study provides a comprehensive analysis of BB in NEC and provides a reference for regional BB management and control. [ABSTRACT FROM AUTHOR]

Over the past four decades, Arctic sea ice coverage has steadily declined. This loss of sea ice has amplified solar radiation and heat absorption from the ocean, exacerbating both polar ice loss and global warming. It has also accelerated changes in sea ice movement, posing safety risks for ship navigation. In recent years, numerical prediction models have dominated the field of sea ice movement prediction. However, these models often rely on extensive data sources, which can be limited in specific time periods or regions, reducing their applicability. This study introduces a novel approach for predicting Arctic sea ice motion within a 10-day window. We employ a Self-Attention ConvLSTM deep learning network based on single-source data, specifically optical flow derived from the Advanced Microwave Scanning Radiometer Earth Observing System 36.5 GHz data, covering the entire Arctic region. Upon verification, our method shows a reduction of 0.80 to 1.18 km in average mean absolute error over a 10-day period when compared to ConvLSTM, demonstrating its improved ability to capture the spatiotemporal correlation of sea ice motion vector fields and provide accurate predictions. [ABSTRACT FROM AUTHOR]

With the continuous development of earth observation technology, space-based synthetic aperture radar (SAR) has become an important source of information for maritime surveillance, and ship classification in SAR images has also become a hot research direction in the field of maritime ship monitoring. In recent years, the remote sensing community has proposed several solutions to the problem of ship object classification in SAR images. However, it is difficult to obtain an adequate amount of labeled SAR samples for training classifiers, which limits the application of machine learning, particularly deep learning methods, in SAR image ship object classification. In contrast, as a real-time automatic tracking system for monitoring ships at sea, a ship automatic identification system (AIS) can provide a large amount of relatively easy-to-obtain labeled ship samples. Therefore, to solve the problem of SAR image ship classification and improve the classification performance of learning models with limited samples, we proposed a SAR image ship classification method based on multiple classifiers ensemble learning (MCEL) and AIS data transfer learning. The core idea of our method is to transfer the MCEL model trained on AIS data to SAR image ship classification, which mainly includes three steps: first, we use the acquired global space-based AIS data to build a dataset for ship object classification models training; then, the ensemble learning model is constructed by combining multiple base classifiers; and finally, the trained classification model is transferred to SAR images for ship type prediction. Experiments show that the proposed method achieves a classification accuracy of 85.00% for the SAR ship classification, which is better than the performance of each base classifier. This proves that AIS data transfer learning can effectively solve the problem of SAR ship classification with limited samples, and has important application value in maritime surveillance. [ABSTRACT FROM AUTHOR]

In this study, we investigate sea state estimation from spaceborne GNSS-R. Due to the complex scattering of electromagnetic waves on the rough sea surface, the neural network approach is adopted to develop an algorithm to derive significant wave height (SWH) from CYGNSS data. Eighty-nine million pieces of CYGNSS data from September to November 2020 and the co-located ECMWF data are employed to train a three-hidden-layer neural network. Ten variables are considered as the input parameters of the neural network. Without the auxiliary of the wind speed, the SWH retrieved using the trained neural network exhibits a bias and an RMSE of −0.13 and 0.59 m with respect to ECMWF data. When considering wind speed as the input, the bias and RMSE were reduced to −0.09 and 0.49 m, respectively. When the incidence angle ranges from 35 ° to 65 ° and the SNR is above 7 dB, the retrieval performance is better than that obtained using other values. The measurements derived from the "Block III" satellite offer worse results than those derived from other satellites. When the distance is considered as an input parameter, the retrieval performances for the areas near the coast are significantly improved. A soft data filter is used to synchronously improve the precision and ensure the desired sample number. The RMSEs of the retrieved SWH are reduced to 0.45 m and 0.41 m from 0.59 m and 0.49 m, and only 16.0% and 14.9% of the samples are removed. The retrieved SWH also shows a clear agreement with the co-located buoy and Jason-3 altimeter data. [ABSTRACT FROM AUTHOR]

This article aims to attempt to increase the number of satellites that can be used for monitoring soil moisture to obtain more precise results using GNSS-IR (Global Navigation Satellite System-Interferometric Reflectometry) technology to estimate soil moisture. We introduce a soil moisture inversion model by using GPS SNR (Signal-to-Noise Ratio) data and propose a novel Robust Kalman Filter soil moisture inversion model based on that. We validate our models on a data set collected at Lamasquère, France. This paper also compares the precision of the Robust Kalman Filter model with the conventional linear regression method and robust regression model in three different scenarios: (1) single-band univariate regression, by using only one observable feature such as frequency, amplitude, or phase; (2) dual-band data fusion univariate regression; and (3) dual-band data fusion multivariate regression. First, the proposed models achieve higher accuracy than the conventional method for single-band univariate regression, especially by using the phase as the input feature. Second, dual-band univariate data fusion achieves higher accuracy than single-band and the result of the Robust Kalman Filter model correlates better to the in situ measurement. Third, multivariate variable fusion improves the accuracy for both models, but the Robust Kalman Filter model achieves better improvement. Overall, the Robust Kalman Filter model shows better results in all the scenarios. [ABSTRACT FROM AUTHOR]

The proposed "Guanlan" ocean science satellite, led by China's Laoshan Laboratory, includes an interferometric radar altimeter (IRA) as a key payload. As an integral part of its development, an airborne IRA experiment was conducted on 6 November 2021, with a flight path of approximately 90 km in the South China Sea. This study investigates the IRA's ability to observe ocean sea surface height (SSH) across scales ranging from meters to mesoscale. The sea surface height anomaly (SSHA) of the IRA is aligned with the SSHA of the AVISO at scales greater than 30 km, but also demonstrates the ability to capture small-scale SSHA changes in two dimensions. We analyzed wavenumber spectra of SSHA obtained from the airborne IRA, ICESat-2, and SARAL/AltiKa satellite for this region. The results show a good agreement in power spectral density (PSD) levels between ICESat-2, SARAL/AltiKa and IRA at scales larger than 30 km. Within the submesoscale range of 1–10 km, the IRA SSHA spectrum exhibits a distinctly negative slope and the lowest energy level. The minimum PSD level of the IRA fell in the range of 10−4–10−3 m2/cycle/km, at scales around 1 km, which is more than an order of magnitude lower than that of ICESat-2, forming a spectral gap that is in agreement with the theoretical expectation. Furthermore, IRA-derived wave direction and significant wave height matched well with the MFWAM wave data. The results of this study underscore the considerable potential of airborne IRA in capturing SSHA across a range of scales, from oceanic waves to submesoscale. [ABSTRACT FROM AUTHOR]

The operation of underwater vehicles in deep waters is a very challenging task. The use of AUVs (Autonomous Underwater Vehicles) is the preferred option for underwater exploration activities. They can be autonomously navigated and controlled in real time underwater, which is only possible with precise spatio-temporal information. Navigation and positioning systems based on LBL (Long-Baseline) or USBL (Ultra-Short-Baseline) systems have their own characteristics, so the choice of system is based on the specific application scenario. However, comparative experiments on AUV navigation and positioning under both systems are rarely conducted, especially in the deep sea. This study describes navigation and positioning experiments on AUVs in deep-sea scenarios and compares the accuracy of the USBL and LBL/SINS (Strap-Down Inertial Navigation System)/DVL (Doppler Velocity Log) modes. In practice, the accuracy of the USBL positioning mode is higher when the AUV is within a 60° observation range below the ship; when the AUV is far away from the ship, the positioning accuracy decreases with increasing range and observation angle, i.e., the positioning error reaches 80 m at 4000 m depth. The navigational accuracy inside and outside the datum array is high when using the LBL/SINS/DVL mode; if the AUV is far from the datum array when climbing to the surface, the LBL cannot provide accurate position calibration while the DVL fails, resulting in large deviations in the SINS results. In summary, the use of multi-sensor combination navigation schemes is beneficial, and accurate position information acquisition should be based on the demand and cost, while other factors should also be comprehensively considered; this paper proposes the use of the LBL/SINS/DVL system scheme. [ABSTRACT FROM AUTHOR]

The retrieval of crop growth status using Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) has become a major area of interest within the field of vegetation remote sensing in recent years. Using only a single GNSS antenna, it is difficult to determine the crop growth status and soil water content (SWC) in vegetation-covered regions due to plenty of multi-path signals. Based on the empirical mode decomposition and the spectrum difference, this study presents an algorithm that can decompose and separate signals reflected by the soil surface or corn canopy. Because the low-roughness soil surface is isotropic while the corn canopy is anisotropic, the signals reflected by the soil surface have a higher proportion of coherent components than those reflected by the corn canopy. The moduli between the retrieved heights and the actual heights (for the same interval from different satellites) have the least variance. In this study, the signals reflected by the soil surface and the corn canopy are separated using the variance of retrieved heights. When the corn grows taller than the GNSS antenna, the vegetation water content (VWC) of the corn leaves becomes the primary factor affecting the direct signal's intensity, as the leaves obstruct the signal. Hence, the VWC of corn leaves can be calculated through the power attenuation of signals. An experiment performed on a plot of land covered with corn shows that, after multi-GPS-satellite fusion, the correlations between the retrieved corn canopy height, leaf VWC, soil water content (SWC), and in situ data reach 0.94, 0.92, and 0.88, respectively. The corresponding root mean square errors are 0.195 m, 0.0055 kg/cm 2 , and 0.0484 cm 3 /cm 3 , respectively. [ABSTRACT FROM AUTHOR]

This research and collaboration work aims at the calibration and validation (Cal/Val) of the Chinese HY-2B satellite altimeter based upon two permanent Cal/Val facilities: (1) the China Altimetry Calibration Cooperation Plan in Qingdao, Bohai Sea and the Wanshan islands, China and (2) the permanent facility for altimetry calibration established by the European Space Agency in Crete, Greece. The HY-2B satellite altimeter and its radiometer have been calibrated and monitored using uniform, standardized procedures, as well as protocols and best practices, and they also built upon trusted and indisputable reference standards at both Cal/Val infrastructures in Europe and China. The HY-2B altimeter is thus monitored in a coordinated, absolute, homogeneous, long-term and worldwide manner. Calibration of altimeters is accomplished by examining satellite observations in open seas against reference measurements. Comparisons are established through precise satellite positioning, water level observations, GPS buoys and reference models (geoid, mean dynamic topography, earth tides, troposphere and ionosphere), all defined at the Cal/Val sites. In this work, the final uncertainty for the altimeter bias will be attributed to several individual sources of uncertainty, coming from observations in water level, atmosphere, absolute positioning, reference surface models, transfer of heights from Cal/Val sites to satellite observations, etc. Through this project, the procedures, protocols and best practices, originally developed in the course of the ESA FRM4ALT project, are updated, upgraded and followed at both Cal/Val facilities in Europe and China. All in all, the HY-2B satellite altimeter observes the sea level quite well and within its specifications. [ABSTRACT FROM AUTHOR]

Passive microwave (PMW) sensors are popularly applied to Earth observations. However, the satellite PMW radiometer data sometimes have non-negligible errors in geolocation. Coastline inflection methods (CIMs) are widely used to improve geolocation errors of PMW images. However, they commonly require accuracy satellite flight parameters, which are difficult to obtain by users. In this study, a simplified coastline inflection method (SCIM) is proposed to correct the geolocation errors without demanding for the satellite flight parameters. SCIM is applied to improve geolocation errors of FengYun-3D (FY-3D) Microwave Radiation Imager (MWRI) brightness temperature images from 2018 and 2019. It reduces the geolocation errors of MWRI images to 0.15 pixels in the along-track and cross-track direction. This means reductions of 75% and 86% of the geolocation errors, respectively. The mean brightness temperature differences between the ascending and descending MWRI images are reduced by 34%, demonstrating the improved geolocation accuracy of SCIM. The corrected images are also used to estimate Arctic sea ice concentration (SIC). By comparing with SICs retrieved from the un-corrected images, the root mean square error (RMSE) and mean absolute error (MAE) of the SICs from the corrected images are reduced from 13.7% to 10.2% and 8.9% to 6.9%, respectively. The mean correlation coefficient (R) increases from 0.91 to 0.95. All these results indicate that SCIM can reduce geolocation errors of satellite-based PMW images significantly. As SCIM is very simple and easy to be applied, it could be a useful method for users of PMW images. [ABSTRACT FROM AUTHOR]

With the development of deep learning techniques, convolutional neural networks (CNN) are increasingly being used in image recognition for marine surveys and underwater object classification. Automatic recognition of targets on side-scan sonar (SSS) images using CNN can improve recognition accuracy and efficiency. However, the vast selection of CNN models makes it challenging to select models for target recognition in SSS images. Therefore, this paper aims to compare different CNN models' prediction accuracy and computational performance comprehensively. First, four traditional CNN models were applied to train and predict the same submarine SSS dataset using both the original model and models with transfer learning methods. Then, we examined and studied the prediction accuracy and computation performance of four CNN models. Results showed that transfer learning enhances the accuracy of all CNN models, with lesser improvements for AlexNet and VGG-16 and greater improvements for GoogleNet and ResNet101. GoogleNet has the highest prediction of accuracy (100% in the train dataset and 94.27% in the test dataset) and good computational difficulty. The findings of this work are useful for future model selection in target recognition in SSS images. [ABSTRACT FROM AUTHOR]

In the process of using InSAR technology to identify active landslides, phenomena such as steep terrain, dense vegetation, and complex clouds may lead to the missed identification of some landslides. In this paper, an active landslide identification method combining InSAR technology and optical satellite remote sensing technology is proposed, and the method is successfully applied to the Three Parallel Rivers Region (TPRR) in the northwest of Yunnan Province, China. The results show that there are 442 landslides identified in the TPRR, and the fault zone is one of the important factors affecting the distribution of landslides in this region. In addition, 70% of the active landslides are distributed within 1 km on both sides of the fault zone. The larger the scale of the landslide, the closer the relationship between landslides and the fault zone. In this identification method, the overall landslide identification rate based on InSAR technology is 51.36%. The combination of Sentinel-1 and ALOS-2 data is beneficial to improve the active landslide identification rate of InSAR. In the northern region with large undulating terrain, shadows and overlaps occur easily. The southern area with gentle terrain is prone to the phenomenon where the scale of landslides is too small. Both phenomena are not conducive to the application of InSAR. Thus, in the central region, with moderate terrain and slope, the identification rate of active landslides based on InSAR is highest. The active landslide identification method proposed in this paper, which combines InSAR and optical satellite remote sensing technology, can integrate the respective advantages of the two technical methods, complement each other's limitations and deficiencies, reduce the missed identification of landslides, and improve the accuracy of active landslide inventory maps. [ABSTRACT FROM AUTHOR]

In the Bering Sea slope, ocean eddies are essential physical processes that carry nutrients to the shelf. The development of the satellite altimeter has facilitated the observation of oceanic eddies. Attention networks are used as the core algorithm for eddy detection to suppress feature responses in irrelevant non-eddy areas, which can address the issue of sample imbalance in high-latitude ocean eddies. Furthermore, data from both the sea surface height (SSH) and geostrophic velocity were employed as model inputs to integrate more eddy-related properties. The results of ocean eddy detection using this method and the dataset allowed more eddies to be detected than with traditional vector geometry-based methods and only SSH-based models. This study also incorporated the results of multiple deep learning models to increase both the overall and single-day eddy detection efficiency. As a result, the algorithms in this paper show that attention networks and geostrophic velocity data are both appropriate for high-latitude ocean eddy identification. This makes a contribution to the further application of deep learning methods to satellite altimetry data. [ABSTRACT FROM AUTHOR]

On 15 January 2022, a violent eruption and tsunami of the Hunga Tonga-Hunga Ha'apai (HTHH) volcano in Tonga, South Pacific, caused widespread international concern. In order to detect the anomalous environmental response caused by the HTHH volcanic eruption based on GNSS ionospheric data, GNSS tropospheric data and GNSS coordinate time series, a new method combining the zenith non-hydrostatic delay difference method and the extreme-point symmetric mode decomposition (ESMD) method, was proposed to detect tropospheric anomalies. The moving interquartile range method and the ESMD method were introduced to detect ionospheric anomalous and coordinate time series anomalies, respectively. The results showed that 9–10 h before the eruption of the Tonga volcano and 11–12 h after the eruption of the Tonga volcano, obvious total electron content (TEC) anomalies occurred in the volcanic eruption center and its northeast and southeast, with the maximum abnormal value of 15 TECU. Significant tropospheric anomalies were observed on the day of the HTHH volcano eruption as well as 1–3 days and 16–17 days after the eruption, and the abnormal intensity was more than 10 times that of normal. The coordinate time series in direction E showed very significant anomalies at approximately 2:45 p.m. on 14 January, at approximately 4:30 a.m.–5:40 a.m. on 15 January, and at approximately 3:45 a.m. on 16 January, with anomalies reaching a maximum of 7–8 times daily. The abnormality in the direction north (N) is not obvious. Very prominent anomalies can be observed in the direction up (U) at approximately 4:30 a.m.–5:40 a.m., with the intensity of the anomalies exceeding the normal by more than 10 times. In this study, GNSS was successfully used to detect the anomalous environmental response during this HTHH volcano eruption. [ABSTRACT FROM AUTHOR]

Deep-learning (DL)-based object detection algorithms can greatly benefit the community at large in fighting fires, advancing climate intelligence, and reducing health complications caused by hazardous smoke particles. Existing DL-based techniques, which are mostly based on convolutional networks, have proven to be effective in wildfire detection. However, there is still room for improvement. First, existing methods tend to have some commercial aspects, with limited publicly available data and models. In addition, studies aiming at the detection of wildfires at the incipient stage are rare. Smoke columns at this stage tend to be small, shallow, and often far from view, with low visibility. This makes finding and labeling enough data to train an efficient deep learning model very challenging. Finally, the inherent locality of convolution operators limits their ability to model long-range correlations between objects in an image. Recently, encoder–decoder transformers have emerged as interesting solutions beyond natural language processing to help capture global dependencies via self- and inter-attention mechanisms. We propose Nemo: a set of evolving, free, and open-source datasets, processed in standard COCO format, and wildfire smoke and fine-grained smoke density detectors, for use by the research community. We adapt Facebook's DEtection TRansformer (DETR) to wildfire detection, which results in a much simpler technique, where the detection does not rely on convolution filters and anchors. Nemo is the first open-source benchmark for wildfire smoke density detection and Transformer-based wildfire smoke detection tailored to the early incipient stage. Two popular object detection algorithms (Faster R-CNN and RetinaNet) are used as alternatives and baselines for extensive evaluation. Our results confirm the superior performance of the transformer-based method in wildfire smoke detection across different object sizes. Moreover, we tested our model with 95 video sequences of wildfire starts from the public HPWREN database. Our model detected 97.9% of the fires in the incipient stage and 80% within 5 min from the start. On average, our model detected wildfire smoke within 3.6 min from the start, outperforming the baselines. [ABSTRACT FROM AUTHOR]

The hybrid interferometric synthetic aperture radar system is a combination of an along-track configuration and cross-track configuration. Based on linear ocean wave theory, an ocean wave inversion algorithm for a hybrid interferometric synthetic aperture radar system is proposed in this work. Using the interferometric synthetic aperture radar images acquired by the TerraSAR-X and TanDEM-X satellites and the interferometric synthetic aperture radar images acquired by an airborne interferometric radar altimeter with a certain degree of squint, the profile of ocean waves and the corresponding orbital velocities were retrieved by combining the new inversion algorithm with the cross-spectra. Meanwhile, key parameters of ocean waves, such as the significant wave height, significant wave orbital velocity, propagation direction, and wavelength of the dominant waves, were also extracted from the ocean wave spectra retrieved in this study. In order to evaluate the reliability of the new inversion algorithm, the retrieved significant wave heights were compared with those provided by the European Centre for Medium-Range Weather Forecasts and measured by a Global Navigation Satellite System buoy. The results showed that for the ocean waves retrieved from the spaceborne hybrid interferometric synthetic aperture radar images, the differences between the retrieved significant wave heights of the four subareas selected in this paper and those provided by European Centre for Medium-Range Weather Forecasts were approximately 0.01, –0.17, –0.55, and –0.37 m, respectively, and for the ocean waves retrieved from the airborne interferometric radar altimeter images, the differences between the retrieved significant wave heights corresponding to the M920 and M3120 images used in this paper and those measured by the Global Navigation Satellite System buoy were approximately –0.05 and –0.09, respectively. Therefore, the method proposed in this work could retrieve the ocean wave spectra well when the velocity bunching had a small influence; however, as the nonlinear influence of the velocity bunching increased, the difference between the significant wave heights retrieved using this method and provided by the European Centre for Medium-Range Weather Forecasts also increased. [ABSTRACT FROM AUTHOR]

In 2007, China discovered a hydrothermal anomaly in the Longqi hydrothermal area of the Southwest Indian Ridge. It was the first seabed hydrothermal area discovered in the ultraslow spreading ocean ridge in the world. Understanding the types of seabed sediments in this area is critical for studying the typical topography and geological characteristics of deep-sea seabed hydrothermal areas. The traditional classification of deep-seabed sediments adopts box sampling or gravity column sampling and identifies the types of seabed sediments through laboratory analysis. However, this classification method has some shortcomings, such as the presence of discrete sampling data points and the failure of full-coverage detection. The geological sampling in deep-sea areas is particularly inefficient. Hence, in this study, the EM122 multibeam sonar data collected in the Longqi hydrothermal area, Southwest Indian Ridge, in April 2019 are used to analyze multibeam backscatter intensity. Considering various errors in the complex deep-sea environment, obtaining backscatter intensity data can truly reflect seabed sediment types. Through unsupervised and supervised classification, the seabed sediment classification in the Longqi hydrothermal area was studied. The results showed that the accuracy of supervised classification is higher than that of unsupervised classification. Thus, unsupervised classification is primarily used for roughly classifying sediment types without on-site geological sampling. Combining the genetic algorithm (GA) and the support vector machine (SVM) neural network, deep-sea sediment types, such as deep-sea clay and calcareous ooze, can be identified rapidly and efficiently. Based on comparative analysis results, the classification accuracy of the GA-SVM neural network is higher than that of the SVM neural network, and it can be effectively applied to the high-precision classification and recognition of deep-sea sediments. In this paper, we demonstrate the fine-scale morphology and surface sediment structure characteristics of the deep-sea seafloor by finely processing high-precision deep-sea multibeam backscatter intensity data. This research can provide accurate seabed topography and sediment data for the exploration of deep-sea hydrothermal resources and the assessment of benthic habitats in deep-sea hydrothermal areas. [ABSTRACT FROM AUTHOR]

The southwest mountainous area of China is one of the areas with the most landslides in the world. In this paper, we used Ya'an City and Garzê Tibetan Autonomous Prefecture in Sichuan Province as the research areas to explore the identification application effects of large-area potential landslides using synthetic aperture radar (SAR) data with different wavelength types (Sentinel-1, ALOS-2), different processing methods (SBAS-InSAR, Stacking-InSAR), and different geological environmental conditions. The results show the following: (1) The effect of identifying landslides with different slope directions is largely affected by the satellite orbit direction; when we identify landslide hazards across a large area, the joint monitoring mode of ascending and descending orbit data is required. (2) The period of monitoring affects the identification effect of potential landslides when landslide identification is carried out in southwestern China; the InSAR monitoring period is recommended to be more than 2 years. (3) In different geological environmental regions, SBAS technology and Stacking technology have their own advantages; Stacking technology identifies more potential landslides, and SBAS technology identifies potential landslides with higher accuracy; (4) the degree of vegetation coverage has a great impact on the landslide identification effect of different SAR data sources. In low-density vegetation coverage areas, the landslide identification result using Sentinel-1 data seems to be better than the result using ALOS-2 data. In high-density vegetation coverage areas, the landslide identification result using ALOS-2 data is better than that using Sentinel-1 data. [ABSTRACT FROM AUTHOR]

The modernization of Global Navigation Satellite System (GNSS) including the transmission of signals on multiple frequencies has greatly promoted the development of the popular PPP (Precise Point Positioning) technique. A key issue of multi-frequency PPP is the handling of the observable-specific signal biases in order to allow for carrier-phase ambiguity resolution (AR). As a result, PPP modeling at a user side in the multi-frequency case varies depending on the definition of the applied phase bias products. In this study, we investigate the positioning performance of GPS L1/L2/L5 and Galileo E1/E5a/E5b/E6 undifferenced ionosphere-float model in the conventional PPP mode and the single-epoch mode using the uncombined code and phase bias products generated at the French CNES (Centre National D'Etudes Spatiales). A series of widelane ambiguities are configured in our multi-frequency PPP functional model instead of forming the classical Melbourne–Wübbena (MW) combination. The best integer equivariant (BIE) estimator is used for the ambiguity resolution in a conventional cascading scheme according to the wavelength of the combined ambiguities for each constellation. Real data collected at IGS stations with a 30-s sampling interval is applied to evaluate the above models. For the conventional kinematic PPP configuration, a significant accuracy improvement of 63% on the east component of the fixed solution is obtained with respect to the ambiguity-float solution. The PPP convergence is accelerated by 17% after the AR. Regarding the single-epoch positioning, an accuracy of 32 and 31 cm for north and east components can be achieved, respectively, (68th percentile) with the instantaneous widelane-ambiguity resolution, which is improved by 13% and 16% compared to multi-frequency code-based or float solution. [ABSTRACT FROM AUTHOR]

The visible-IR scanning radiometer (VIRR) of FY1-C/D meteorological satellites consists of 10 bands with 4 different focal plane assemblies (FPAs). However, there are significant band-to-band registration (BBR) errors between different bands, which cannot be compensated for by a simple shift in the along-scan direction. A rigorous BBR frame was proposed to analyze the sources of misregistration in the whisk-broom camera. According to theory, the 45° scanning mirror introduces tangent function style misregistration in the along-track direction and secant function style misregistration in the across-track direction between different bands if the bands are not in the same optical axis. As proven by the experiments of both FY-1C and FY-1D, the image rotation caused by the 45° scanning mirrors plays a major role in the misregistration. However, misregistration between different FPAs does not strictly adhere to this theory. Therefore, a polynomial-based co-registration method was proposed to model the BBR errors for the VIRR. To achieve 0.1 pixel accuracy, a fourth-degree polynomial was used for BBR in the along-scan direction, and a fifth-degree polynomial was used for the along-track direction. For the reflective bands, the root-mean-square errors (RMSEs) of misregistration could be improved from 3 pixels to 0.11 pixels. Limited by matching accuracy, the RMSEs of misregistration between thermal bands and reflective bands were approximately 0.2 to 0.4 pixels, depending on the signal-to-noise ratio. [ABSTRACT FROM AUTHOR]

Global navigation satellite system interferometric reflectometry (GNSS-IR) represents an extra method to detect snow depth for climate research and water cycle managing. However, using a single frequency of GNSS-IR for snow depth retrieval is often found to be challenging when attempting to achieve a high spatial and temporal sensitivity. To evaluate both the capability of the GNSS-IR snow depth retrieved by the multi-GNSS system and multi-frequency from signal-to-noise ratio (SNR) data, the accuracy of snow depth retrieval by different frequency signals from the multi-GNSS system is analyzed, and a joint retrieval is carried out by combining the multi-GNSS system retrieval results. The SNR data of the global positioning system (GPS), global orbit navigation satellite system (GLONASS), Galileo satellite navigation system (Galileo), and BeiDou navigation satellite system (BDS) from the P387 station of the U.S. Plate Boundary Observatory (PBO) are analyzed. A Lomb–Scargle periodogram (LSP) spectrum analysis is used to compare the difference in reflector height between the snow-free and snow surfaces in order to retrieve the snow depth, which is compared with the PBO snow depth. First, the different frequency retrieval results of the multi-GNSS system are analyzed. Then, the retrieval accuracy of the different GNSS systems is analyzed through multi-frequency mean fusion. Finally, the joint retrieval accuracy of the multi-GNSS system is analyzed through mean fusion. The experimental shows that the retrieval results of different frequencies of the multi-GNSS system have a strong correlation with the PBO snow depth, and that the accuracy is better than 10 cm. The multi-frequency mean fusion of different GNSS systems can effectively improve the retrieval accuracy, which is better than 7 cm. The joint retrieval accuracy of the multi-GNSS system is further improved, with a correlation coefficient (R) between the retrieval snow depth and the PBO snow depth of 0.99, and the accuracy is better than 3 cm. Therefore, using multi-GNSS and multi-frequency data to retrieve the snow depth has a good accuracy and feasibility. [ABSTRACT FROM AUTHOR]

Understory vegetation plays an important role in the structure and function of forest ecosystems. Light detection and ranging (LiDAR) can provide understory information in the form of either point cloud or full-waveform data. Point cloud data have a remarkable ability to represent the three-dimensional structures of vegetation, while full-waveform data contain more detailed information on the interactions between laser pulses and vegetation; both types have been widely used to estimate various forest canopy structural parameters, including leaf area index (LAI). Here, we present a new method for quantifying understory LAI in a temperate forest by combining the advantages of both types of LiDAR data. To achieve this, we first estimated the vertical distribution of the gap probability using point cloud data to automatically determine the height boundary between overstory and understory vegetation at the plot level. We then deconvolved the full-waveform data to remove the blurring effect caused by the system pulse to restore the vertical resolution of the LiDAR system. Subsequently, we decomposed the deconvolved data and integrated the plot-level boundary height to differentiate the waveform components returned from the overstory, understory, and soil layers. Finally, we modified the basic LiDAR equations introducing understory leaf spectral information to quantify the understory LAI. Our results, which were validated against ground-based measurements, show that the new method produced a good estimation of the understory LAI with an R2 of 0.54 and a root-mean-square error (RMSE) of 0.21. Our study demonstrates that the understory LAI can be successfully quantified through the combined use of point cloud and full-waveform LiDAR data. [ABSTRACT FROM AUTHOR]

Sea surface height can be measured with the delay between reflected and direct global navigation satellite system (GNSS) signals. The arrival time of a feature point, such as the waveform peak, the peak of the derivative waveform, and the fraction of the peak waveform is not the true arrival time of the specular signal; there is a bias between them. This paper aims to analyze and calibrate the bias to improve the accuracy of sea surface height measured by using the reflected signals of GPS CA, Galileo E1b and BeiDou B1I. First, the influencing factors of the delay bias, including the elevation angle, receiver height, wind speed, pseudorandom noise (PRN) code of GPS CA, Galileo E1b and BeiDou B1I, and the down-looking antenna pattern are explored based on the Z-V model. The results show that (1) with increasing elevation angle, receiver height, and wind speed, the delay bias tends to decrease; (2) the impact of the PRN code is uncoupled from the elevation angle, receiver height, and wind speed, so the delay biases of Galileo E1b and BeiDou B1I can be derived from that of GPS CA by multiplication by the constants 0.32 and 0.54, respectively; and (3) the influence of the down-looking antenna pattern on the delay bias is lower than 1 m, which is less than that of other factors; hence, the effect of the down-looking antenna pattern is ignored in this paper. Second, an analytical model and a neural network are proposed based on the assumption that the influence of all factors on the delay bias are uncoupled and coupled, respectively, to calibrate the delay bias. The results of the simulation and experiment show that compared to the meter-level bias before the calibration, the calibrated bias decreases the decimeter level. Based on the fact that the specular points of several satellites are visible to the down-looking antenna, the multi-observation method is proposed to calibrate the bias for the case of unknown wind speed, and the same calibration results can be obtained when the proper combination of satellites is selected. [ABSTRACT FROM AUTHOR]