Gaofen-3 Research Articles

Snowy Dove: An open-sourcetoolkit for pre- processing of Chinese Gaofen series data.

Since the launch of the Chinese High-resolution Earth Observation System (CHEOS) program, China has strengthened its research and development in the field of satellite remote sensing. A large number of sensors has been or will be launched, providing very large data streams which all require processing of the engineering data, as provided by the instruments, to physical data which will be used for further processing and interpretation. To handle such large data streams we developed a one-click batch pre-processing toolkit for CHEOS remote sensing data as described in this paper. In this toolkit, IDL language and environment are used as the primary program combined with other programming languages developed in this research. In this paper, we first describe the Gaofen (GF) series data used in this research and then introduce the function design and realization of this one-click batch pre-processing toolkit. Some examples will be presented to illustrate the application of the toolkit to data from several CHEOS satellites.

Unveiling Illumination Variations During a Lunar Eclipse: Multi-Wavelength Spaceborne Observations of the January 21, 2019 Event

Space-based observations of the total lunar eclipse on 21 January 2019 were conducted using the geostationary Earth-orbiting satellite Gaofen-4 (GF-4). This study represents a pioneering effort to address the observational gap in full-disk lunar eclipse photometry from space. With its high resolution and ability to capture the entire lunar disk, GF-4 enabled both quantitative and qualitative analyses of the variations in lunar brightness, as well as spectra and color changes, across two spatial dimensions, from the whole lunar disk to resolved regions. Our results indicate that before the totality phase of the lunar eclipse, the irradiance of the Moon diminishes to below approximately 0.19% of that of the uneclipsed Moon. Additionally, we observed an increase in lunar brightness at the initial entry into the penumbra. This phenomenon is attributed to the opposition effect, providing scientific evidence for this unexpected behavior. To investigate detailed spectral variations, specific calibration sites, including the Chang’E-3 landing site, MS-2 in Mare Serenitatis, and the Apollo 16 highlands, were analyzed. Notably, the red-to-blue ratio dropped below 1 near the umbra, contradicting the common perception that the Moon appears red during lunar eclipses. The red/blue ratio images reveal that as the Moon enters Earth’s umbra, it does not simply turn red; instead, a blue-banded ring appears at the boundary due to ozone absorption and the lunar surface composition. These findings significantly enhance our understanding of atmospheric effects on lunar eclipses and provide crucial reference information for the future modeling of lunar eclipse radiation, promoting the integration of remote sensing science with astronomy.

Preliminary Global NO2 Retrieval from EMI-II Onboard GF5B/DQ1 and Comparison to TROPOMI

The Environmental Trace Gases Monitoring Instrument (EMI-II) onboard the Chinese GaoFen-5B (GF5B) and DaQi-1 (DQ1) satellites is the successor of the previous EMI onboard the Chinese GaoFen-5 (GF5) satellite, and has a higher spatial resolution and a better signal-to-noise ratio. The GF5B and DQ1 were launched in September 2021 and April 2022, respectively. As part of China’s ultraviolet-visible hyperspectral satellite instrument series, the EMI-II aims to conduct network observations of pollution gases globally in the morning and early afternoon. In this study, NO2 data were retrieved from the EMI-II payloads on the GF5B and DQ1 satellites using the Differential Optical Absorption Spectroscopy (DOAS) algorithm. The two satellites were consistently compared, and the results showed strong consistency on various spatial and temporal scales (R2 > 0.8). In four representative regions worldwide, NO2 data from the EMI-II exhibited good spatial consistency with those from the TROPOMI. The correlation coefficient (R2) of the total vertical column density (VCD) between the EMI-II and TROPOMI exceeded 0.85, and that of the tropospheric NO2 VCD exceeded 0.57. Compared with single-satellite observations, the dual-satellite network of the GF5B and DQ1 can effectively increase the observation frequency. On a daily scale, dual-satellite observations can reduce the impact of cloud coverage by 6–8% compared to single-satellite observations, and there are two valid observations of nearly 50% of the world’s regions. Additionally, the differences between the two satellites can reflect the NO2 diurnal variations, which demonstrates the potential for studying pollutant gas diurnal variations.

Object-oriented U-GCN for open-pit mining extraction from high spatial resolution remote-sensing images of complex scenes

ABSTRACT Precise extraction of open-pit mines is crucial for resource management and ecological and environmental dynamic monitoring. Current methods for extracting open-pit mines encounter challenges such as low accuracy and difficulty detecting complex scenes of open-pit mining. To address these issues, this paper proposes an object-oriented intelligent extraction method for complex mining scenes using Gaofen-2(GF-2) high-resolution remote-sensing images, which expresses the pixel-level features at the object level and utilizes the unique feature propagation and aggregation capabilities of the graph structure for the extraction of the mines. First, an object-oriented feature expression strategy is introduced to express multi-level pixel-level features as object-level features by constructing objects with appropriate multi-resolution segmentation parameters. This approach effectively reduces the impact of isolated pixel noise and outliers on classification results. Second, this paper proposes a U-GCN-based open-pit mining extraction method that combines the powerful multi-level feature extraction capabilities of U-GCN to propagate and aggregate information in a graph structure, effectively modelling spatial relationships between different objects. This method achieves high-precision extraction of open-pit mining areas. In experiments conducted on two study areas of varying scales, the F1 scores for open-pit mine extraction reached 93.32% and 83.06%. Comparative experiments demonstrate that the proposed object-oriented U-GCN method performs superiorly in terms of accuracy, stability and robustness across mining scenes with different levels of complexity. The proposed open-pit mine extraction method offers new insights and methodologies for current extraction practices.

Retrieval and Comparison of Multi-Satellite Polar Ozone Data from the EMI Series Instruments

The Environmental Trace Gases Monitoring Instrument (EMI) series are second-generation Chinese spectrometers on board the GaoFen-5 (GF-5) and DaQi-1 (DQ-1) satellites. In this study, a comparative analysis of EMI series data was conducted to determine the daily trend of ozone concentration changes owing to different transit times and to improve the overall quality and reliability of EMI series datasets. The daily EMI total ozone column (TOC) obtained using the Differential Optical Absorption Spectroscopy (DOAS) method were compared to vertical column density (VCD) gathered by the TROPOspheric Monitoring Instrument (TROPOMI). The results from October to November 2023 indicated a fine correlation (R = 0.98) between the daily EMI series data and a fine correlation (R ≥ 0.95) and spatial distribution closely resembling that of the TROPOMI TOCs. Furthermore, the EMI series data fusion results were highly correlated with TROPOMI TOCs (R = 0.99). Since the EMI series instruments had two different overpass times and the volume of available data at same pixel was increased by approximately three-fold, the temporal and spatial resolution was improved a lot. The results indicated that, compared to a single sensor, the EMI series DOAS TOCs generated more accurate and stable global TOC results and also enabled looking at the changes in the intraday TOCs. These outcomes highlight the potential of the EMI instruments for reliably monitoring the ozone variations in polar regions.

Ship Lock Extraction from High-Resolution Remote Sensing Images Based on Fuzzy Theory and Prior Knowledge

As crucial water conservancy projects, ship locks play a key role in flood control, shipping, water resource allocation, and promoting regional economic development, making them an indispensable part of the modern water transportation system. Utilizing satellite remote sensing for lock extraction can significantly reduce manual workload and costs, assist in the daily dynamic maintenance of lock hubs, and provide more comprehensive data support for the construction and management of water transport infrastructure. In this context, this paper proposes a new method for ship lock object extraction. Leveraging fuzzy theory and prior knowledge of locks, the extraction of lock objects is achieved from Gaofen-1 (GF-1) high-resolution remote sensing images. The experimental results demonstrate that the proposed algorithm can effectively extract small lock objects in remote sensing images, achieving an average extraction accuracy of 80.9% in the study area.

Influence of sea surface waves on numerical modeling of an oil spill: Revisit of symphony wheel accident

The greatest purpose of this study is to analyze the importance of surface waves on the hindcasting of the oil spill through the Symphony wheel accident in the Qingdao coastal waters. During the accident period, a total of four synthetic aperture radar (SAR) images by Gaofen-3 (GF-3) were acquired from 2 to 19 May 2021. The hindcasting of two sea surface dynamics, namely currents and waves, is carried out using a coupled marine numeric model. This model, known as the finite-volume community ocean model-simulating waves nearshore (FVCOM-SWAVE), employs a triangular grid. Simulated significant wave height (SWH) is validated against remotely sensed product by the Haiyang-2B (HY-2B) altimeter on April 2021 yields a root mean square error (RMSE) of 0.38, a correlation coefficient (COR) of 0.78, and a scatter index (SI) of 0.34. Subsequently, Stokes drift estimated by waves are included to hindcasting oil spills using the oil particle-tracing method. The bias of the spatial coverage (SAR minus simulations) of an algorithm called the constant false alarm rate (CFAR) is −73.92 km2 with Stokes drift, which is significantly less than the 55.45 km2 coverage without Stokes drift. Moreover, compared with model-simulated oil spills, the bias of the geographic location at the center point with Stokes drift is 8.18 km, which is less than the 12.95 km bias without Stokes drift. These results demonstrate that Stokes drift needs to be included in the prediction of oil spills.

RSU-Net: A new method for fine classification of corn residue coverage in black soil area using Chinese GF-1B PMS image

Corn residue coverage (CRC) plays an important role in soil protection, with respect to both the restoration of soil fertility and the reduction of soil erosion, particularly in black soil are-as. Fine classification of areas with high and low coverage of crop residue is essential for accu-rately quantifying the soil protection level; it is insufficient to focus solely on whether the crop residue is covered. Remote sensing images from Chinese GaoFen-1 (GF-1) B, which can be acquired with a high spatial resolution of 2 m and a revisit period of 4d, have potential for the fine classification of high and low coverage of corn residues (HLCCR). To improve the efficiency and accuracy of the fine classification of HLCCR, a deep neural network RSU-Net was proposed in this study. RSU-Net combines residual units with spatial pyramid pooling technology to enhance the ability to extract the features of different coverage types from GF-1 B images. Comparison ex-periments were conducted to validate the effectiveness of RSU-Net and compare it with the classical U-Net. The results show that the overall accuracy, kappa, and mean intersection over union of RSU-Net in fine classification of HLCCR were 86.29%, 0.75, and 0.6844, respectively, and those of U-Net were 84.61%, 0.72, and 0.6474, respectively. Compared with the classical U-Net, the classification accuracy of RSU-Net was improved. These results suggest that RSU-Net has great po-tential for fine classification of HLCCR.

Comparative analysis of multispectral data between GF-1 WFV4 and GF-6 WFV sensors

ABSTRACT In recent years, China has launched a number of Gaofen (GF) series earth observation satellites. It is crucial to understand the relationship between the data of the GF series of satellite sensors for the selection of sensor images for scientific research. Taking the same-day transit image pairs of three regions as the study data, this paper compares the consistency of the Top of Atmosphere (TOA) reflectance of GF-1 WFV4 and GF-6 WFV sensors by using the TOA mean comparison method, and discusses the differences in water body and vegetation extraction. The results indicate that the satellite signal intensity in different land cover types and areas differs significantly. In the bare soil-dominated region, GF-1 WFV4 has a larger signal strength than GF-6 WFV, while in the vegetation-dominated region, it turns out to be just the opposite. The difference between the two sensors is mainly related to the difference in the spectral response function and the radiometric resolution of the two satellites. In addition, the determination coefficient (R2) of the corresponding bands of the two sensors is all greater than 0.90, indicating that the two sensors have strong linear correlation and good complementarity.

Estimation of Picea Schrenkiana Canopy Density at Sub-Compartment Scale by Integration of Optical and Radar Satellite Images

This study proposes a novel approach to estimate canopy density in Picea Schrenkiana var. Tianschanica forest sub-compartments by integrating optical and radar satellite data. This effort is aimed at enhancing methodologies for forest resource surveys and monitoring, particularly vital for the sustainable development of semi-arid mountainous areas with fragile ecological environments. The study area is the West Tianshan Mountain Nature Reserve in Xinjiang, which is characterized by its unique dominant tree species, Picea Schrenkiana. A total of 411 characteristic factors were extracted from Gaofen-2 (GF-2) sub-meter optical satellite imagery, Gaofen-3 (GF-3) multi-polarization synthetic aperture radar satellite imagery, and digital elevation model (DEM) data. Consequently, 17 characteristic parameters were selected based on their correlation with canopy density data to construct an estimation model. Three distinct models were developed, including a multiple stepwise regression model (a linear approach), a Back Propagation (BP) neural network model (a neural network-based method), and a Cubist model (a decision tree-based technique). The results indicate that combining optical and radar image characteristics significantly enhances accuracy, with an Average Absolute Percentage Precision (AAPP) value improvement in estimation accuracy from 76.50% (with optical image) and 78.50% (with radar image) to 78.66% (with both). Of the three models, the BP neural network model achieved the highest overall accuracy (79.19%). At the sub-component scale, the BP neural network model demonstrated superior accuracy in low canopy density estimation (75.37%), whereas the Cubist model, leveraging radar image characteristics, excelled in medium density estimations (87.46%). Notably, the integrated Cubist model combining optical and radar data achieved the highest accuracy for high canopy density estimation (89.17%). This study highlights the effectiveness of integrating optical and radar data for precise canopy density assessment, contributing significantly to ecological resource monitoring methodologies and environmental assessments.

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.

Radiometric Calibration of the Near-Infrared Bands of GF-5-02/DPC for Water Vapor Retrieval

The GaoFen (GF)-5-02 satellite is one of the new generations of hyperspectral observation satellites launched by China in 2021. The directional polarimetric camera (DPC) is an optical sensor onboard the GF-5-02 satellite. The precipitable water vapor (PWV) is a key detection parameter of DPC. However, the existing PWV data developed using DPC data have significant errors due to the lack of the timely calibration of the two bands (865, 910 nm) of DPC used for PWV retrieval. In order to acquire DPC PWV data with smaller errors, a calibration method is developed for these two bands. The method consists of two parts: (1) calibrate the 865 nm band of the DPC using the cross-calibration method, (2) calibrate the 910 nm band of the DPC according to the calibrated 865 nm band of the DPC. This method effectively addresses the issue of the absence of a calibration method for the water vapor absorption band (910 nm) of the DPC. Regardless of whether AERONET PWV data or SuomiNET PWV data are used as the reference data, the accuracy of the DPC PWV data developed using calibrated DPC data is significantly superior to that of the DPC PWV data retrieved using data before recalibration. This means that the calibration method for the NIR bands of the DPC can effectively enhance the quality of DPC PWV data.

Urban lakes change extraction using time series GaoFen-1 satellite imagery

Abstract. Urban lakes serve an indispensable role in maintaining the ecological balance of cities, ensuring flood safety, and providing recreational spaces for tourism. With the development of human activities and economic, the extent of urban lakes are inevitably influenced. Currently, the ability to detect detailed temporal changes in urban lake areas using high resolution data still has limitations. This study proposed a novel method by combining time series Gaofen-1 (GF-1) remote sensing data and random forest machine learning algorithm to explore the urban lakes change Zhushan Lake located in Wuhan. The research conducted the extraction of surface water for Zhushan Lake and its surrounding pit-ponds from 2013 to 2020. And then, a quantitative analysis of the characteristics and driving factors of lake changes is conducted. We find that (1) the accuracy of surface water extraction using the random forest classification method consistently exceeded 96%. The Kappa coefficient ranges from a minimum of 0.86 to a maximum of 0.99. (2) A noticeable decline was observed in the water areas of Zhushan Lake and its surrounding pit-ponds, predominantly along the northwestern shoreline and in the eastern pond regions. This decline is primarily attributed to pressures from building construction. The methodology proposed in this study is suitable for the area management of lakes in urban areas.

A Deep Learning Classification Scheme for PolSAR Image Based on Polarimetric Features

Polarimetric features extracted from polarimetric synthetic aperture radar (PolSAR) images contain abundant back-scattering information about objects. Utilizing this information for PolSAR image classification can improve accuracy and enhance object monitoring. In this paper, a deep learning classification method based on polarimetric channel power features for PolSAR is proposed. The distinctive characteristic of this method is that the polarimetric features input into the deep learning network are the power values of polarimetric channels and contain complete polarimetric information. The other two input data schemes are designed to compare the proposed method. The neural network can utilize the extracted polarimetric features to classify images, and the classification accuracy analysis is employed to compare the strengths and weaknesses of the power-based scheme. It is worth mentioning that the polarized characteristics of the data input scheme mentioned in this article have been derived through rigorous mathematical deduction, and each polarimetric feature has a clear physical meaning. By testing different data input schemes on the Gaofen-3 (GF-3) PolSAR image, the experimental results show that the method proposed in this article outperforms existing methods and can improve the accuracy of classification to a certain extent, validating the effectiveness of this method in large-scale area classification.

Long-term assessment and analysis of the radiometric quality of standard data products for Chinese Gaofen-1/2/6/7 optical remote sensing satellites

The first Chinese Gaofen (GF) remote sensing satellite was launched in August 2013 and has been in orbit for more than 10 years, providing a rich variety of image product data for remote sensing applications in various industries, with other remote sensing satellites of the GF series. To ensure the reliability of the information generated via remote sensing applications, all remote sensing images must undergo systematic quality evaluation. The quality of GF satellites is evaluated almost comprehensively during in-orbit testing at the early stage after launch; however, the remaining evaluations of product quality including geometric or radiometric calibrations are performed only annually. Thus, long-term systematic image quality evaluations are lacking, and the quality level and long-term change trends of the image products are unknown. These deficits have introduced uncertainty in the application of these products. This study aimed to establish a radiometric quality evaluation index framework for GF satellites to comprehensively measure the radiometric quality of GF image products. Multiple types of sample data, such as uniform field, knife-edge target, and radiometric calibration field data with different reflectance features, were acquired globally. Four core indexes, namely, the relative radiometric correction accuracy (RRCA), absolute radiometric correction accuracy (ARCA), modulation transfer function (MTF), and signal-to-noise ratio (SNR), were used to analyze and evaluate the long-term radiometric quality of the images from four GF optical remote sensing satellites: GF1, GF2, GF6, and GF7. The four GF satellites had high SNR and RRCA but low dynamic MTF and variable ARCA for different reflectance features. For example, the average SNRs for the four satellites reached 42.66, 43.91, 44.33, and 47.46 dB; the average RRCA (@root-mean-square errors) values reached 1.6%, 1.0%, 1.6%, and 0.1%; the average MTF values (at Nyquist frequency) of the panchromatic bands were only 0.15, 0.12, 0.07 and 0.04; and the difference in ARCA for different reflectance targets was approximately 10%. In the long time series, the SNRs for each sensor of the four GF satellites were high and stable for high-reflectance features, whereas the RCA and SNR values for low-reflectance features and the MTF of the sensors fluctuated in a cyclic manner. Meanwhile, large dynamic fluctuations and episodic image quality problems occurred at certain local time points, which must be considered during practical applications. The findings presented here will lay a foundation for the in-depth application of GF images.

Twin Satellites HY-1C/D Reveal the Local Details of Astronomical Tide Flooding into the Qiantang River, China

This article extracts the Qiantang River tidal bore, analyzing the water environment characteristics in front of the tidal line of the Qiantang River tidal bore and behind it. The Qiantang River tidal bore Index (QRI) was established using HY-1C, HY-1D, and Gao Fen-1 wide field-of-view (GF-1 WFV) satellite data to precisely determine the location and details of the Qiantang River tidal bore. Comparative analyses of the changes on the two sides of the Qiantang River tidal bore were conducted. The results indicate the following: (1) QRI enhances the visibility of tidal bore lines, accentuating their contrast with the surrounding river water, resulting in a more vivid character. QRI proves to be an effective extraction method, with potential applicability to similar tidal lines in different regions. (2) Observable roughness changes occur at the tidal bore location, with smoother surface textures observed in front of the tidal line compared to those behind it. There is a discernible increase in suspended sediment concentration (SSC) as the tidal bore passes through. (3) This study reveals the mechanism of water environment change induced by the Qiantang River tidal bore, emphasizing its significance in promoting vertical water body exchange as well as scouring the bottom sediments. This effect increases SSC and surface roughness.

Application of LA-UNet network model in remote sensing classification of urban green space

The accurate identification and monitoring of urban green space is of great significance in urban planning and ecological management. In view of the complex background of urban green space, the traditional remote sensing classification technology is prone to the problem of misalignment and adhesion. Taking Yuhua District of Changsha City as the research area and Gaofen-2 (GF-2) remote sensing image as the data source, we proposed a remote sensing classification method for urban green space based on the LA-UNet model, which was based on the UNet model. We introduced the DWTCA channel attention mechanism module to improve the attention of the network to green space information, and used the CARAFE module to up sample the extracted features to achieve accurate classification of trees, shrubs and other land types in the complex background of the city. The results showed that the LA-UNet model had the best classification effect of urban green space when using standard false color remote sensing images. The overall accuracy and mean intersection over union were 96.3% and 90.9%, which were 2.8% and 6.1% higher than the UNet model, respectively. In the Potsdam public dataset, the overall accuracy and mean intersection over union of the LA-UNet model were also better than those of the UNet model, which increased by 0.9% and 1.8%, respectively, indicating that the LA-UNet model had good robustness and versatility. In summary, the proposed LA-UNet model could effectively alleviate the problems of misalignment and adhesion of urban green space, with advantages in the remote sensing classification of urban green space. The improved LA-UNet model had a smaller parameter volume than the UNet model, which could effectively improve the classification accuracy of urban green space. This study would provide a methodological reference for the accurate classification and understanding the spatial distribution of urban green space.

A New High-Resolution Rural Built-Up Land Extraction Method Based on Artificial Surface Index with Short-Wave Infrared Downscaling

The complexity of surface characteristics in rural areas poses challenges for accurate extraction of built-up areas from remote sensing images. The Artificial Surface Index (ASI) emerged as a novel and accurate built-up land index. However, the absence of short-wave infrared (SWIR) bands in most high-resolution (HR) images restricts the application of index-based methods in rural built-up land extraction. This paper presents a rapid extraction method for high-resolution built-up land in rural areas based on ASI. Through the downscaling techniques of random forest (RF) regression, high-resolution SWIR bands were generated. They were then combined with visible and near-infrared (VNIR) bands to compute ASI on GaoFen-2 (GF-2) images (called ASIGF). Furthermore, a red roof index (RRI) was designed to reduce the probability of misclassifying built-up land with bare soil. The results demonstrated that SWIR downscaling effectively compensates for multispectral information absence in HR imagery and expands the applicability of index-based methods to HR remote sensing data. Compared with five other indices (UI, BFLEI, NDBI, BCI, and PISI), the combination of ASI and RRI achieved the optimal performance in built-up land enhancement and bare land suppression, particularly showcasing superior performance in rural built-up land extraction.

DaliWS: A High-Resolution Dataset with Precise Annotations for Water Segmentation in Synthetic Aperture Radar Images

The frequent occurrence of global flood disasters leads to millions of people falling into poverty each year, which poses immense pressure on governments and hinders social development. Therefore, providing more data support for flood disaster detection is of paramount importance. To facilitate the development of water body detection algorithms, we create the DaliWS dataset for water segmentation, which contains abundant pixel-level annotations, and consists of high spatial resolution SAR images collected from the GaoFen-3 (GF-3) satellite. For comprehensive analysis, extensive experiments are conducted on the DaliWS dataset to explore the performance of the state-of-the-art segmentation models, including FCN, SegNeXt, U-Net, and DeeplabV3+, and investigate the impact of different polarization modes on water segmentation. Additionally, to probe the generalization of our dataset, we further evaluate the models trained with the DaliWS dataset, on publicly available water segmentation datasets. Through detailed analysis of the experimental results, we establish a valuable benchmark and provide usage guidelines for future researchers working with the DaliWS dataset. The experimental results demonstrate the F1 scores of FCN, SegNeXt, U-Net, and DeeplabV3+ on the dual-polarization data of DaliWS dataset reach to 90.361%, 90.192%, 92.110%, and 91.199%, respectively, and these four models trained using the DaliWS dataset exhibit excellent generalization performance on the public dataset, which further confirms the research value of our dataset.

Soil Classification Mapping Using a Combination of Semi-Supervised Classification and Stacking Learning (SSC-SL)

In digital soil mapping, machine learning models have been widely applied. However, the accuracy of machine learning models can be limited by the use of a single model and a small number of soil samples. This study introduces a novel method, semi-supervised classification combined with stacking learning (SSC-SL), to enhance soil classification mapping in hilly and low-mountain areas of Northern Jurong City, Jiangsu Province, China. This study incorporated Gaofen-2 (GF-2) remote sensing imagery along with its associated remote sensing indices, the ALOS Digital Elevation Model (DEM) and their derived topographic factors, and soil parent material data in its modelling process. We first used three base learners, Ranger, Rpart, and XGBoost, to construct the SL model. In addition, we employed the fuzzy c-means clustering algorithm (FCM) to construct a clustering map. To fully leverage the information from a multitude of environmental variables, understand the distribution of data, and enhance the effectiveness of the classification, we selected unlabelled samples near the boundaries of the patches on the clustering map. The SSC-SL model demonstrated superior stability and performance, with optimal accuracy at a 0.9 confidence level, achieving an overall accuracy of 0.77 and a kappa coefficient of 0.73. These metrics exceeded those of the highest performing base learner (Ranger model) by 10.4% and 12.3%, respectively, and they outperformed the least effective base learner (Rpart model) by 27.3% and 32.9%. It notably improves the spatial distribution accuracy of soil types. Key environmental variables influencing soil type distribution include soil parent material (SPM), land use (LU), the multi-resolution valley bottom flatness index (MRVBF), and Elevation (Ele). In conclusion, the SSC-SL model offers a novel and effective approach for enhancing the predictive accuracy of soil classification mapping.