Sentinel-2 Images Research Articles

Use of HANTS algorithm in Sentinel-2 images for interpolation of missing data: a case-study for soy agricultural areas in periods of growth and harvesting

Abstract. Information about the current state of an agricultural culture can be fundamental in decision making, varying from quantity of fertilizers to be used in an area to estimating the productivity of a harvest in a period to establish credits payment. Satellite images with reasonable spatial resolution like Sentinel-2 are a powerful tool to visualize agricultural areas production, especially with the application of multispectral indexes like the Normalized Difference Vegetation Index (NDVI). Yet, due to the limitation of temporal resolution of satellites and periods with large cloudy weather can make the analysis difficult. This study proposed the use of the harmonic analysis time series algorithm (HANTS) to aid the incompleteness of time series obtained by the available satellite images in a case study of soy culture in Brazil. By applying the algorithm, a harmonic interpolation is obtained to produce a full daily NDVI time series from the beginning to end of the study period, facilitating further analysis to obtain metrics of interest. In this study, time series with gaps on specified periods are created based on full time series, then interpolated and compared with the real ones based on root mean squared error (RMSE) to access its accuracy.

Coherence and Intensity Statistical Mapping of Earthquake Affected Areas using Sentinel-1 Images

Abstract. The January 1, 2024, Noto Peninsula earthquake left many devastated areas. Right after the event, aerial photography provided the initial assessment but covered only specific areas. Satellite images with optical sensors that are freely available are limited by spatial resolution and oftentimes covered by cloud. The Sentinel-1 Synthetic Aperture Radar (SAR) due to its all-weather sensing capability proved to be useful in mapping the extent of the affected areas. This study mapped Wajima City where earthquake aftereffects were severe as reported using coherence and intensity mapping. Pre- and post-event images were used to rapidly identify affected areas. Based on the result, taking the coherence of one pre- and one post-event images can give an impression of wide earthquake damage. The coherence of two post-event images compared with two pre-event images provides a better estimate of affected areas. Taking multiple pre- and post- event images improved the map because the large variation of the intensity is minimized through time-series averaging. Using a simple RGB composite, the affected areas were mapped. Threshold mapping was also used to extract collapsed buildings/houses using a training dataset. The mean coherence difference and mean intensity difference between pre- and post-event images were the two variables used. The computed minimum threshold for the mean coherence difference was 0.35. However, the average not the minimum of the mean intensity difference of 3.0 was used as initial value. Using mean coherence difference>=0.35 gave a high accurate prediction of collapsed buildings at 73.47% but also has false identification of 30.95%. To negate the overprediction, the mean intensity difference<=4.0 was integrated. The result lowered the overprediction to 20.75% but also lowered the accuracy to 64.97%. This assessment was only conducted in the fire-razed morning market in Wajima City.

Identifying Damaged Areas Caused by the Noto Peninsula 2024 Earthquake Using Change Detection from Optical Satellite Images

Abstract. Strategic response during disasters requires the swiftness of the flow of data and information. Simple change detections can be used from Optical Satellite Images to achieve fast results that can be used for analysis and planning. In response to the earthquake that devastated the Noto Peninsula of Japan on New Year’s Day, news reports and early optical satellite images were utilized to identify areas that were damaged by the earthquake. The changes in the reflectance values and indicators, such as NDBI and NDVI were analyzed to identify the areas with damages such as collapsed buildings and landslides. From the after-event satellite images of PlanetScope and Sentinel-2, three areas were identified as severely affected based on news reports namely, Wajima City, Niemachi, and Noto Town. To verify the results of the analysis, aerial images provided by the Geographical Survey Institute of Japan, and Google Earth Street View Images were used. Through change detection methods, the locations of damaged areas were pinpointed using Optical Satellite Images after the earthquake incident.

Precursory crater contraction associated with the 2017 eruption of Shinmoe-dake volcano (Japan) detected by PALSAR-2 and Sentinel-1 InSAR

The time series of PALSAR-2 and Sentinel-1 images reveal inflation at the volcanic flank and contraction at the crater for approximately 5 months before the 2017 eruption of Shinmoe-dake volcano, Japan. While the observation of inflation at the volcano’s flank is ubiquitous, few studies have reported crater contraction preceding an eruption. The flank inflation stopped after the 2017 eruption, while the contraction at the crater continued until the 2018 eruption. We found that a pipe-shaped deformation source above sea level best fits the observation preceding the 2017 eruption. Suppose the flux of ejected materials constrains the conduit radius during the previous 2011 eruption. In that case, the amount of deformation of the pipe-shaped deformation source, whether open or closed at its top, is too large to be realistic. Although constraining the conduit radius from the eruption flux overestimates the pressure change of the pipe-shaped deformation source, water-saturated fractures along the volcanic conduit could extend the effective conduit radius of the pressure source. We propose one potential scenario for the mechanism of the crater contraction preceding volcanic eruptions based on the combination of compaction due to cooling by ambient groundwater and material withdrawal within the conduit. The groundwater inflows from the ambient aquifer through cracks in the porous conduit wall, which are generated by conduit expansion during the magma ascent. Decoupling from the conduit wall due to a decrease in volume of the material promotes material instability and crater contraction. The interaction between the groundwater and the magma triggers the 2017 eruption of Shinmoe-dake volcano, as previous studies have reported.Graphical

Improving airborne laser scanning-based species-specific forest volume estimation using sentinel-2 time series

ABSTRACT Species-specific timber volume estimates are required to support forest planning and conservation. We evaluated whether additional predictors from a Sentinel-2 time series can improve airborne laser scanning (ALS)-based estimation of species-specific timber volumes. Furthermore, we determined the satellite image dates that provided the greatest improvement in accuracy. The Sentinel-2 time series was constructed to cover 1 March–30 November time-period, with a focus on late spring and early summer. The estimation was done using the k Most Similar Neighbor method and predictors extracted from the ALS data and Sentinel-2 images. Our best model included both ALS and Sentinel-2 time-series predictors, and the relative root mean square error (RMSE) values for pine, spruce and deciduous timber volumes were 40.8%, 57.0% and 51.3%, respectively (mean 49.7%). All deciduous trees were treated as one species. When bands from an individual image were used instead of the time series, the best result was obtained with an image from September where the respective relative RMSE values were 42.2% (deciduous), 58.4% (pine) and 60.6% (spruce) with a mean value of 53.7%. A fusion of a Sentinel-2 time-series and ALS data can improve species-specific estimation results compared to the use of individual Sentinel-2 images or ALS only.

Identify and map coastal aquaculture ponds and their drainage and impoundment dynamics

Sustainable management of coastal aquaculture ponds could achieve win-win between food and economic benefits and ecological conservation including waterbird. In this study, 5,790 Harmonized Landsat and Sentinel-2 images from July 2021 to June 2022 and 498 Sentinel-1 images from July 2021, August 2021, and June 2022 as supplementary data were collected to calculate multiple water indices. Based on Otsu algorithm to distinguish between water and non-water region and Savitzky-Golay filtering to optimize time series, coastal aquaculture ponds were identified using the SNIC. Furthermore, their drainage and impoundment phases were determined using the Dynamic Time Warping-Kmeans++ method. Finally, a new 30-m resolution dataset at the national scale of China was generated with an overall accuracy greater than 90 % for both the pond map and the drainage and impoundment phases. Our observations revealed that the total area was 7,919.53 km2, with the largest pond area in Shandong Province. Among the coastal aquaculture ponds, 27.95 % were seasonal aquaculture ponds, 70.32 % were yearlong aquaculture ponds, and 1.49 % were abandoned aquaculture ponds. Drainage start dates, end dates, and durations were calculated based on abrupt changes in the water proportion time series. Drainage start dates were concentrated from September to December, while drainage end dates were from January to April. Drainage durations of coastal aquaculture ponds ranged from two weeks to six months, with Shanghai Municipality having the longest drainage durations and Taiwan Province having the shortest drainage durations. The findings could provide scientific support for modifying the drainage and impoundment phases of coastal aquaculture ponds to achieve the win–win goal of improving economic development and protecting waterbirds or improving offshore water quality.

Effects of model-overfit on model-assisted forest inventory in boreal forests with remote sensing data

Abstract While remote sensing can be an effective tool in building a forest inventory, field measurements and model fitting can be both expensive and challenging. One strategy to reduce forest inventory costs is to leverage forest inventory models fitted to a different population (external models), although the effectiveness of external models is poorly understood. One concern is that models may predict well to the sample data, but poorly to the population—which is termed ‘overfitting’. The effect of overfit may be especially problematic in attempts to predict for a different population (a forest area not covered by any sample plots). Assessing overfit is difficult and its consequence for estimation are not well understood, especially in the context of prediction using external models. This study assesses how overfitting affects model-assisted forest inventory estimation when using internal and external models. We used field and remotely sensed data (Sentinel-2 images and airborne laser scanning data) from two forest areas in Finland. We evaluated four modeling approaches: ordinary least square regression (OLS), random forest, k-nearest neighbors, and gaussian process regression. Both analytical and bootstrap variance estimators were used to evaluate model-assisted estimation performance. Internal models, especially OLS, were the most affected by model overfitting, leading to bias in the population means and underestimation of variance. Estimates using external models provided unbiased means and realistic intervals except in the case of deliberate excessive overfitting. The bootstrap variance estimator was found to be more robust to overfit than the analytical variance estimator for the internal model, but was not helpful for the external model. Internal models should be parsimonious to generalize well to the population and avoid bias. The bootstrap estimator of variance is recommended for internal models, especially if there is concern about model overfitting.

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

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

Spatiotemporal Relationship Between Land Subsidence and Ecological Environmental Quality in Shenfu Mining Area, Loess Plateau, China

The exploitation of coal resources has caused problems such as ground deformation, affecting the ecological environment. Spatiotemporal varying characteristics between land subsidence and ecological environmental quality (EEQ) are an important research hotspot. Using the SBAS-InSAR method, 64 Sentinel-1 images were utilized to monitor land subsidence in the Shenfu mining area, one of China’s largest coal source regions. And the remote sensing ecological index (RSEI) was used to monitor and evaluate EEQ of the Shenfu mining area. Global and local spatial autocorrelation methods were used to assess the spatial aggregation degree and change patterns over time. Spatial Econometric Models were employed to explore the impacts of land subsidence on EEQ. The results showed the following: (1) The average RSEI values in the Shenfu mining area were 0.531, 0.488, and 0.523 in 2016, 2018, and 2020, respectively; there was a slight downward trend in EEQ. The permanent scatter (PS) point deformation rate ranged from −353.40 mm/year to +246.24 mm/year, with average deformation rates of 0.1642, 0.2181, and 0.2490 mm/year, respectively. (2) There was a significant correlation and spatial agglomeration effect between land surface subsidence and EEQ. Low–high, high–low, and low–low clusters were the main types of relationships, indicating that land subsidence primarily has a negative spatial impact on the ecological environment. (3) The relationship between land subsidence and EEQ varied spatially in the Shenfu mining area at 500 × 500 grid units. This research can provide scientific guidance for disaster prevention and sustainable development in mining areas by considering long-term differences in ecological environmental quality and its correlation with land subsidence.

MSMTRIU-Net: Deep Learning-Based Method for Identifying Rice Cultivation Areas Using Multi-Source and Multi-Temporal Remote Sensing Images

Identifying rice cultivation areas in a timely and accurate manner holds great significance in comprehending the overall distribution pattern of rice and formulating agricultural policies. The remote sensing observation technique provides a convenient means to monitor the distribution of rice cultivation areas on a large scale. Single-source or single-temporal remote sensing images are often used in many studies, which makes the information of rice in different types of images and different growth stages hard to be utilized, leading to unsatisfactory identification results. This paper presents a rice cultivation area identification method based on a deep learning model using multi-source and multi-temporal remote sensing images. Specifically, a U-Net based model is employed to identify the rice planting areas using both the Landsat-8 optical dataset and Sentinel-1 Polarimetric Synthetic Aperture Radar (PolSAR) dataset; to take full into account of the spectral reflectance traits and polarimetric scattering traits of rice in different periods, multiple image features from multi-temporal Landsat-8 and Sentinel-1 images are fed into the network to train the model. The experimental results on China’s Sanjiang Plain demonstrate the high classification precisions of the proposed Multi-Source and Multi-Temporal Rice Identification U-Net (MSMTRIU-NET) and that inputting more information from multi-source and multi-temporal images into the network can indeed improve the classification performance; further, the classification map exhibits greater continuity, and the demarcations between rice cultivation regions and surrounding environments reflect reality more accurately.

Vegetative Index Intercalibration Between PlanetScope and Sentinel-2 Through a SkySat Classification in the Context of “Riserva San Massimo” Rice Farm in Northern Italy

Rice farming in Italy accounts for about 50% of the EU’s rice area and production. Precision agriculture has entered the scene to enhance sustainability, cut pollution, and ensure food security. Various studies have used remote sensing tools like satellites and drones for multispectral imaging. While Sentinel-2 is highly regarded for precision agriculture, it falls short for specific applications, like at the “Riserva San Massimo” (Gropello Cairoli, Lombardia, Northern Italy) rice farm, where irregularly shaped crops need higher resolution and frequent revisits to deal with cloud cover. A prior study that compared Sentinel-2 and the higher-resolution PlanetScope constellation for vegetative indices found a seasonal miscalibration in the Normalized Difference Vegetation Index (NDVI) and in the Normalized Difference Red Edge Index (NDRE). Dr. Agr. G.N. Rognoni, a seasoned agronomist working with this farm, stresses the importance of studying the radiometric intercalibration between the PlanetScope and Sentinel-2 vegetative indices to leverage the knowledge gained from Sentinel-2 for him to apply variable rate application (VRA). A high-resolution SkySat image, taken almost simultaneously with a pair of Sentinel-2 and PlanetScope images, offered a chance to examine if the irregular distribution of vegetation and barren land within rice fields might be a factor in the observed miscalibration. Using an unsupervised pixel-based image classification technique on SkySat imagery, it is feasible to split rice into two subclasses and intercalibrate them separately. The results indicated that combining histograms and agronomists’ expertise could confirm SkySat classification. Moreover, the uneven spatial distribution of rice does not affect the seasonal miscalibration object of past studies, which can be adjusted using the methods described here, even with images taken four days apart: the first method emphasizes accuracy using linear regression, histogram shifting, and histogram matching; whereas the second method is faster and utilizes only histogram matching.

Detection of Complex Formations in an Inland Lake from Sentinel-2 Images Using Atmospheric Corrections and a Fully Connected Deep Neural Network

The detection of complex formations, initially suspected to be oil spills, is investigated using atmospherically corrected multispectral satellite images and deep learning techniques. Several formations have been detected in an inland lake in Northern Greece. Four atmospheric corrections (ACOLITE, iCOR, Polymer, and C2RCC) that are specifically designed for water applications are examined and implemented on Sentinel-2 multispectral satellite images to eliminate the influence of the atmosphere. Out of the four algorithms, iCOR and ACOLITE are able to depict the formations sufficiently; however, the latter is chosen for further processing due to fewer uncertainties in the depiction of these formations as anomalies across the multispectral range. Furthermore, a number of formations are annotated at the pixel level for the 10 m bands (red, green, blue, and NIR), and a deep neural network (DNN) is trained and validated. Our results show that the four-band configuration provides the best model for the detection of these complex formations. Despite not being necessarily related to oil spills, studying these formations is crucial for environmental monitoring, pollution detection, and the advancement of remote sensing techniques.

Flooded area detection and mapping from Sentinel-1 imagery. Complementary approaches and comparative performance evaluation

ABSTRACT The current study assesses the performance of several machine learning (ML) and deep learning (DL) models for detecting and mapping floods using Sentinel-1 SAR imagery. Three distinct approaches were used: pixel classification, object-based image analysis and object instance segmentation. The ML models are Random Forest, and Support Vector Machine and the DL models are U-NET, DeepLabV3 and Mask RCNN. Five different case studies were selected to test the models’ scalability. These areas are in Romania (Prut River, at the border between Romania, the Republic of Moldova and Ukraine, Timiș River, and Râul Negru River), the United States of America (Missouri River) and Australia (Broughton Creek). For each flood, five Sentinel-1 images were used, four collected before the flood and one collected after the flood. The intensity images were stacked and scaled in the range of the intensity thresholds associated with water and non-water so that all the case studies have the same margins for intensity. Samples of water, vegetation, agricultural fields, and bare soil were collected only from the Prut River case study and used in the training process. Out of all models, the U-Net model returned the highest accuracy with a value for Intersect over Union of 0.763 for a tile size of 128x128 pixels.

Monitoring Salinity in Inner Mongolian Lakes Based on Sentinel-2 Images and Machine Learning

Salinity is an essential parameter for evaluating water quality and plays a crucial role in maintaining the stability of lake ecosystems, particularly in arid and semi-arid climates. Salinity responds to changes in climate and human activity, with significant impacts on water quality and ecosystem services. In this study, Sentinel-2A/B Multi-Spectral Instrument (MSI) images and quasi-synchronous field data were utilized to estimate lake salinity using machine learning approaches (i.e., XGB, CNN, DNN, and RFR). Atmospheric correction for MSI images was tested using six processors (ACOLITE, C2RCC, POLYMER, MUMM, iCOR, and Sen2Cor). The most accurate model and atmospheric correction method were found to be the extreme gradient boosting tree combined with the ACOLITE correction algorithm. These were used to develop a salinity model (N = 70, mean absolute percentage error = 9.95%) and applied to eight lakes in Inner Mongolia from 2016 to 2024. Seasonal and interannual variations were explored, along with an examination of potential drivers of salinity changes over time. Average salinities in the autumn and spring were higher than in the summer. The highest salinities were observed in the lake centers and tended to be consistent and homogeneous. Interannual trends in salinity were evident in several lakes, influenced by evaporation and precipitation. Climate factors were the primary drivers of interannual salinity trends in most lakes.

Estimation of Urban Tree Chlorophyll Content and Leaf Area Index Using Sentinel-2 Images and 3D Radiative Transfer Model Inversion

Urban trees play an important role in mitigating effects of climate change and provide essential ecosystem services. However, the urban environment can stress trees, requiring the use of effective monitoring methods to assess their health and functionality. The objective of this study, which focused on four deciduous tree species in Rennes, France, was to evaluate the ability of hybrid inversion models to estimate leaf chlorophyll content (LCC), leaf area index (LAI), and canopy chlorophyll content (CCC) of urban trees using eight Sentinel-2 (S2) images acquired in 2021. Simulations were performed using the 3D radiative transfer model DART, and the hybrid inversion models were developed using machine-learning regression algorithms (random forest (RF) and gaussian process regression). Model performance was assessed using in situ measurements, and relations between satellite data and in situ measurements were investigated using spatial allocation (SA) methods at the pixel and tree scales. The influence of including environment features (EFs) as model inputs was also assessed. The results indicated that random forest models that included EFs and used the pixel-scale SA method were the most accurate with R2 values of 0.33, 0.29, and 0.46 for LCC, LAI, and CCC, respectively, with notable variability among species.

Above ground biomass estimation in the upper Blue Nile basin forests, North-Western Ethiopia

Forest ecosystems play a decisive role in the global climatic condition, as well as, provides a wide range of societal benefits, including fuel-wood, tourism, and ecosystem services are considered as one of the major sources of livelihood for the local people in the upper Blue Nile Basin. Therefore, rapid and accurate estimation of forest biomass is crucial for greatly reducing the uncertainty in carbon stock assessments, and for designing strategic forest management plans. Because, above-ground biomass (AGB) estimation is important in determining the management, environmental, and economic roles of forests in the Blue Nile basin. The study was aimed at estimating above-ground biomass in the Upper Blue Nile Basin forests by integrating field-measured data with predictors from Sentinel-2 image. The relationship between measured AGB and sentinel-2 derived vegetation indices and biophysical parameters showed a good correlation result (r value ranging from 0.67 to 0.74). A stepwise regression analysis was carried out in order to develop AGB estimation model by identifying the most important variable. The result demonstrated that, green normalized difference vegetation index, leaf area index, fraction of absorbed photosynthetic active radiation and fractional vegetation cover achieved good performance in predicting AGB with R2 value > 0.5. AGB was estimated with a coefficient of determination (R2) of 0.59 adjusted R2 of 0.618 and root mean square error of (RMSE) 38.36 t/ha in comparison to field observations. The maximum AGB value of 268.32 t/ha was estimated in the Alemsaga natural forest, which is a highly protected dense forest stand from any entrance and disturbance. Generally, integrating field data with optical remote sensing data provides more reliable result for AGB estimation. Moreover, it is also recommended to employ RADAR and LiDAR remote sensing data products together in order to attain more precise estimate results of AGB with great potential for forest resource monitoring and management.

TSAE-UNet: A Novel Network for Multi-Scene and Multi-Temporal Water Body Detection Based on Spatiotemporal Feature Extraction

The application of remote sensing technology in water body detection has become increasingly widespread, offering significant value for environmental monitoring, hydrological research, and disaster early warning. However, the existing methods face challenges in multi-scene and multi-temporal water body detection, including the diverse variations in water body shapes and sizes that complicate detection; the complexity of land cover types, which easily leads to false positives and missed detections; the high cost of acquiring high-resolution images, limiting long-term applications; and the lack of effective handling of multi-temporal data, making it difficult to capture the dynamic changes in water bodies. To address these challenges, this study proposes a novel network for multi-scene and multi-temporal water body detection based on spatiotemporal feature extraction, named TSAE-UNet. TSAE-UNet integrates convolutional neural networks (CNN), depthwise separable convolutions, ConvLSTM, and attention mechanisms, significantly improving the accuracy and robustness of water body detection by capturing multi-scale features and establishing long-term dependencies. The Otsu method was employed to quickly process Sentinel-1A and Sentinel-2 images, generating a high-quality training dataset. In the first experiment, five rectangular areas of approximately 37.5 km2 each were selected to validate the water body detection performance of the TSAE-UNet model across different scenes. The second experiment focused on Jining City, Shandong Province, China, analyzing the monthly water body changes from 2020 to 2022 and the quarterly changes in 2022. The experimental results demonstrate that TSAE-UNet excels in multi-scene and long-term water body detection, achieving a precision of 0.989, a recall of 0.983, an F1 score of 0.986, and an IoU of 0.974, significantly outperforming FCN, PSPNet, DeepLabV3+, ADCNN, and MECNet.

Ensemble Machine-Learning-Based Framework for Estimating Surface Soil Moisture Using Sentinel-1/2 Data: A Case Study of an Arid Oasis in China

Soil moisture (SM) is a critical parameter in Earth’s water cycle, significantly impacting hydrological, agricultural, and meteorological research fields. The challenge of estimating surface soil moisture from synthetic aperture radar (SAR) data is compounded by the influence of vegetation coverage. This study focuses on the Weigan River and Kuche River Delta Oasis in Xinjiang, employing high-resolution Sentinel-1 and Sentinel-2 images in conjunction with a modified Water Cloud Model (WCM) and the grayscale co-occurrence matrix (GLCM) for feature parameter extraction. A soil moisture inversion method based on stacked ensemble learning is proposed, which integrates random forest, CatBoost, and LightGBM. The findings underscore the feasibility of using multi-source remote sensing data for oasis moisture inversion in arid regions. However, soil moisture content estimates tend to be overestimated above 10% and underestimated below 5%. The CatBoost model achieved the highest accuracy (R2 = 0.827, RMSE = 0.014 g/g) using the top 16 feature parameter groups. Additionally, the R2 values for Stacking1 and Stacking2 models saw increases of 0.008 and 0.016, respectively. Thus, integrating multi-source remote sensing data with Stacking models offers valuable support and reference for large-scale estimation of surface soil moisture content in arid oasis areas.

Mapping Italian high-altitude ponds.

Permanent and temporary ponds are considered peculiar ecosystems which provide important ecosystem functions, services, supporting biodiversity on small and large scales. Pond's conservation status is globally critical. Moreover, their ecological functioning and conservation status is frequently overlooked, because of the habitat small size, their seasonal occurrence and their unique appearance. While a certain attention is given to Mediterranean Temporary Ponds and, in general, to low altitude ponds, the ecological importance of high-altitude ponds is critically unrecognized, especially in the Italian peninsula. The main aim of this research is to create the first georeferenced checklist of Italian high-altitude ponds. In order to achieve this goal, we integrated spectral, spatial characteristics, and morphological operations based on Sentinel-1 and Sentinel-2 image data using the Google Earth Engine (GEE). Overall, 2156 ponds were identified: 62% (n = 1343) in the Alps and 38% (n = 813) in the Apennines. The highest number of ponds was detected in Central Alps (n = 642), followed by Western Alps (n = 479), Central Apennines (n = 412), Eastern Alps (n = 222), Southern Apennines (n = 216) and Northern Apennines (n = 185). For what concerns the Alps, the average altitude was estimated in 2428 m a.s.l., while in the Apennines the average altitude was estimated in 784 m a.s.l. The total area covered from ponds has been estimated in 4.258.640 m2, with a mean of 1716 m2. Ponds were described as 20% temporary (n = 445) and 80% permanent (n = 1711). Considering the land use, 83% (n = 1797) of ponds were described as "natural" and 17% (n = 359) as "anthropized". Identification and georeferentiation of high-altitude ponds are primary actions to the application of management plans and this research could be considered the first step towards the safeguard of these threatened ecosystems.

Super-resolution cropland mapping with Sentinel-2 images based on a self-training learning network

ABSTRACT The Sentinel-2 image is widely used for cropland mapping, but the limitation of its 10 m spatial resolution may significantly impact the accuracy of result in areas characterized by severe cropland fragmentation. To address the issue, this letter proposes a super-resolution cropland mapping model for Sentinel-2 images. The proposed model improves the spatial resolution of Sentinel-2 images to 2.5 m through self-training learning with the Swin Transformer for Image Restoration (SwinIR) model without needing fine-resolution training samples. Then, the random forest classification is applied to map cropland from the super-solved 2.5 m resolution image. The proposed model was assessed in the Jianghan Plain, China, and the results show that the super-resolution images produce cropland maps with higher accuracy than the original Sentinel-2 images. Comparing cropland mapping results before and after super-resolution, the overall accuracy improves from 0.82 to 0.89, while the commission error and omission error decrease from 0.09 to 0.05 and 0.17 to 0.11, respectively. This method addresses the shortage of training samples collection and improves the accuracy of cropland mapping, contributing to more reliable agricultural remote sensing applications.