Remote Sensing Applications-Society and Environment最新文献

{"title":"Cloud gap filling for continuous normalized difference snow index snow cover using the dynamic seasonally recurrent snow depletion pattern over a mountainous watershed","authors":"C.D. Woodruff ,&nbsp;R.J. Qualls ,&nbsp;K.S. Humes","doi":"10.1016/j.rsase.2026.102062","DOIUrl":"10.1016/j.rsase.2026.102062","url":null,"abstract":"<div><div>Cloud cover in remotely sensed snow cover is a persistent issue limiting accurate observation of snow cover. Cloud gap filling (CGF) seeks to infill cloud covered pixels, yet methods have historically focused on binary snow detection. Continuous normalized differenced snow index (NDSI) snow cover provides more information about snow cover than binary detection and is a valuable variable for modeling snowmelt runoff. Our goal was to use the dynamic seasonally recurrent snow depletion pattern to produce cloud free continuous NDSI snow cover for the period of 2000-2020 in the Upper Snake River Basin, Wyoming. We test two approaches to CGF for continuous NDSI snow cover, a one-to-one approach where one model is derived for each NDSI snow cover value, and a one-to-many approach where one model is derived to CGF all NDSI snow cover values. We then compared CGF results with the Moderate Resolution Imaging Spectroradiometer (MODIS) CGF product. We demonstrate the dynamic seasonally recurrent snow depletion pattern produces continuous daily NDSI snow cover with 96.23% accuracy over the NDSI snow cover range of 10-50 as calculated against independent cloud free (&lt;10% cloud covered area) images from 2018 to 2020. Our model estimates of snow coverage were up to 10.8% lower than the MODIS CGF product. Accurate daily imagery of snow extent is essential for snowmelt modelers to forecast runoff and manage water resources because up to 1/6th of the world's population relies on water from snowmelt.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 102062"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147957105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"High-resolution machine-learning reconstruction and spatio-temporal analysis of tropospheric OMI NO2 over Bangladesh (2013–2023)","authors":"Mahbuba Hasnat Swa ,&nbsp;Sadit Bihongo Malitha ,&nbsp;A.M. Sarwaruddin Chowdhury","doi":"10.1016/j.rsase.2026.102029","DOIUrl":"10.1016/j.rsase.2026.102029","url":null,"abstract":"<div><div>Nitrogen dioxide (NO₂) pollution poses a severe threat to public health and environmental quality in Bangladesh, yet nationwide monitoring is severely constrained by a sparse ground network. This study developed the first long-term (2013–2023), high-resolution (0.01° × 0.01°, ∼1 km) monthly tropospheric NO₂ dataset over Bangladesh by downscaling OMI Level-3 retrievals using an Extreme Gradient Boosting model trained on 16 geophysical and anthropogenic predictors. Five-fold cross-validation showed excellent performance (R = 0.952, RMSE = 0.464 × 10¹⁵ molecules cm⁻²). The model-based gap-filling increased spatial and temporal coverage of OMI data from 52.84% to 100%, enabling reliable NO₂ estimates even during the cloudy monsoon season. The resulting gap-filled high-resolution dataset revealed persistent NO₂ hotspots in Dhaka and Rajshahi divisions, with 11-year mean values reaching 4.5 × 10¹⁵ molecules cm⁻². Strong seasonal cycles were observed, with winter peaks (up to 5.74 × 10¹⁵ molecules cm⁻²) driven by shallow boundary layers and stagnant conditions, and monsoon minima (&lt;0.7 × 10¹⁵ molecules cm⁻²). Mean predicted OMI NO₂ concentrations fluctuated between 2.23 × 10¹⁵ and 2.81 × 10¹⁵ molecules/cm² during 2013–2023, with the highest value and largest pollution area observed in 2021. Validation against ground stations yielded moderate divisional correlations (R = 0.50–0.61 in Dhaka and Barisal). Local SHAP analysis of the extreme December 2021 hotspot in central Dhaka (12.42 × 10¹⁵ molecules cm⁻²) identified population density, anthropogenic NO_x emissions, and tropospheric total column NO₂ as the features with the largest contributions to the model prediction. This high-resolution dataset provides critical input for exposure assessment, emission inventory refinement, and targeted air-quality policy in one of the world's most polluted and data-scarce regions.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 102029"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147748677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of National Development Policies on Agricultural Land Use Dynamics in Chichaoua, Morocco: A Two-Decade Remote Sensing Analysis","authors":"Ikram El Hazdour ,&nbsp;Michel Le Page ,&nbsp;Lionel Jarlan ,&nbsp;Marielle Montginoul ,&nbsp;Lahoucine Hanich","doi":"10.1016/j.rsase.2026.102034","DOIUrl":"10.1016/j.rsase.2026.102034","url":null,"abstract":"<div><div>Agricultural development policies can boost productivity and profits but also lead to adverse consequences when natural resources are exploited unsustainably, particularly in systems vulnerable to drought and climate change.</div><div>This study provides a novel, comparative upstream-downstream assessment of land-use evolution in Morocco's semi-arid Chichaoua region under national development policies, particularly the Green Moroccan Plan, by integrating multi-source remote sensing. Chichaoua offers a peculiar case where newly-cultivated land upstream of a historic irrigated area facilitates a straightforward evaluation of policy-related land and water use changes. The upstream zone, previously unexploited, relies on boreholes, whereas the downstream area has historically depended on natural springs for traditional irrigation. Utilising Sentinel-2 and Landsat imagery and ground data from 2023, a Random Forest model was employed for crop-classification over 23-years (∼2000-2023), and land-use transitions were analysed through transition matrices. Total irrigated area doubled, increasing from ∼2871 to ∼6569 ha, but with stark contrasts between the zones.</div><div>Upstream areas expanded by +4000 ha, featuring 22% cash-crops and 41% olive trees, while downstream zones witnessed a 40% decline in olive cultivation. National policies offering grants and improving land and water access are correlated with shifts toward cash-crop-focused farming upstream, potentially fuelling productivity gains. However, this transformation appears to be associated with reduced natural spring availability, potentially undermining traditional downstream agriculture. This dual impact highlights both the benefits of increased agricultural productivity and the challenges of resource depletion and threats to traditional farming. The findings reveal critical trade-offs between agricultural gains and environmental risks in vulnerable regions.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 102034"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147750203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"UAV RGB imagery and YOLO11 instance segmentation for gully erosion mapping on tailings dam","authors":"Mariana S. Loo,&nbsp;Claudia E. Durango","doi":"10.1016/j.rsase.2026.101998","DOIUrl":"10.1016/j.rsase.2026.101998","url":null,"abstract":"<div><div>Tailings dam slopes are vulnerable to surface erosion processes that progressively reduce structural stability and increase environmental risk. Among these processes, gully formation represents a critical mechanism of channelized degradation that requires systematic monitoring. This study presents a UAV-based remote sensing workflow integrating high-resolution RGB orthomosaics with YOLO11 instance segmentation for automated delineation of individual gully features on engineered dam slopes. Unlike pixel-wise approaches, the method generates instance-level polygons that enable spatial quantification of erosion extent and direct integration into GIS-based slope-stability and inspection frameworks. The workflow supports rapid, repeatable erosion surveillance and contributes to risk-informed monitoring and preventive maintenance strategies in tailings dam management.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 101998"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613002","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating LUCAS in-situ and Sentinel data for grass height and vigour of vegetation in Europe","authors":"Daniel Gruschwitz ,&nbsp;Mattia Rossi ,&nbsp;Julien Morel ,&nbsp;Marijn van der Velde ,&nbsp;Raphaël d'Andrimont ,&nbsp;Momchil Yordanov ,&nbsp;Tobias Ullmann","doi":"10.1016/j.rsase.2026.102001","DOIUrl":"10.1016/j.rsase.2026.102001","url":null,"abstract":"<div><div>Quantifying grassland biomass is essential for monitoring forage availability in livestock-based agricultural systems and for supporting sustainable land management and policy decisions across Europe. Remote sensing offers extensive capabilities for monitoring grasslands, but the lack of adequate <em>in-situ</em> data for training and validation constrains the evaluation of regression models at a continental scale. The dedicated grassland module in the 2022/23 rollout of the Land Use/Cover Area frame Survey (LUCAS) constitutes a unique, large-scale and geographically comprehensive data source covering the European Union (EU). This study investigates the suitability of the LUCAS variables average grass height and vigour of vegetation for remote sensing-based analysis by conducting machine learning regression and classification on more than 4000 transects with a cloud-free Sentinel-2 scene available. The EU-wide grass height Support Vector Regression model, based on Sentinel-2 bands and vegetation indices, delivers a root mean square error (RMSE) of approximately 19 cm and an R² of 0.25. Predicting the height of tall grassland presents greater challenges due to increased saturation of vegetation indices and the onset of plant senescence. The vigour of vegetation – a visual assessment on an ordinal scale – only allows for classification analysis that leads to frequent misclassification despite class aggregation. In conclusion, the average grass height is the preferred variable within the LUCAS database for biomass-related and remote sensing-based modelling. Rigorous filtering using ground images and metadata is strongly recommended to account for the environmental and management heterogeneity across the EU.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 102001"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613003","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CLOSDI: A novel spectral index for cloud shadow detection in Sentinel-2 imagery using NDVI and EVI2","authors":"Adrián Cal","doi":"10.1016/j.rsase.2026.101990","DOIUrl":"10.1016/j.rsase.2026.101990","url":null,"abstract":"<div><div>Cloud shadows constitute a significant source of uncertainty in optical remote sensing, thereby impairing the accuracy of spectral indices, land cover classifications, and time-series analyses derived from Sentinel-2 imagery. Although the Sentinel-2 Level-2A Scene Classification Layer (SCL) offers information on cloud shadows, its efficacy is frequently hindered by high omission rates and inconsistent detection. This study presents a novel spectral index, namely, the Cloud Shadow Detection Index (CLOSDI), which is specifically designed to enhance cloud shadow detection within agricultural and grassland landscapes.</div><div>CLOSDI is derived from the differential sensitivity of the Normalized Difference Vegetation Index (NDVI) and the two-band Enhanced Vegetation Index (EVI2) to shadow conditions and can be computed directly from RED and NIR reflectance using a closed-form formulation. The index was evaluated using 1231 high-quality image patches from the CloudSEN12 dataset, which are restricted to temperate and tropical grassland biomes. An optimal cutoff threshold of 34.0 was identified through an 80/20 training–testing split by maximizing the median intersection over union (IoU).</div><div>On the test set, CLOSDI substantially outperformed the Sentinel-2 SCL-based shadow mask. The median performance metrics were: recall of 75.6 versus 19.6, F1-score of 63.4 versus 28.6, IoU of 46.4 versus 16.7, and balanced overall accuracy of 80.3 versus 57.6 for CLOSDI and SCL, respectively. A Wilcoxon signed-rank test confirmed that the improvement in IoU was statistically significant (p = 1.402 × 10⁻¹⁹). Despite using only two spectral bands, CLOSDI achieved a performance comparable to state-of-the-art cloud shadow detection algorithms, while requiring substantially lower computational complexity and eliminating the need for model training.</div><div>These findings illustrate that the CLOSDI offers a straightforward, physically interpretable, and computationally efficient approach for cloud shadow detection within vegetation-dominated landscapes. Its ease of implementation and robust performance render it particularly appropriate for large-scale agricultural monitoring, vegetation time-series analysis, and operational remote sensing workflows.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 101990"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-model deep learning framework for urban waste detection in sub-Saharan Africa using Sentinel-2 imagery","authors":"Sachit Desa,&nbsp;Angela Harris,&nbsp;Jonathan J. Huck,&nbsp;Matt D. Tomkins","doi":"10.1016/j.rsase.2026.101985","DOIUrl":"10.1016/j.rsase.2026.101985","url":null,"abstract":"<div><div>Unmanaged and informal waste accumulations remain unmapped across many rapidly urbanizing cities in sub-Saharan Africa. Current approaches often depend on curated training datasets and commercial imagery, which limits their applicability in data-scarce regions. This study addresses this gap by developing a deep learning–based framework to detect formal and informal waste sites across four urban areas in sub-Saharan Africa using open-access Sentinel-2 imagery. The framework's performance was evaluated over four cities—Accra, Abidjan, Dakar, and Lagos. The network integrates semantic segmentation models (HRNet and DeepLabv3+), a classification algorithm (XGBoost), and a texture- and index-based filtering module. Ground-truth validation was carried out using high-resolution imagery and manual annotation over an area covering approximately 10% of the urban extent in each city. The network achieved robust performance, with recall ranging from 0.66 to 0.74, precision from 0.64 to 0.85, F1-score from 0.65 to 0.77, and intersection-over-union (IoU) from 0.48 to 0.64. The detected sites ranged in size from 18.9 m² to over 416,800 m², with most sites located along urban peripheries, road corridors, or near vegetation and water bodies. The resulting outputs represent the most comprehensive mapping to date of waste accumulations across these four urban areas. The proposed framework demonstrates the potential of open-access satellite imagery and machine learning tools to monitor land-use change and track evolving spatial dynamics in rapidly urbanizing and morphologically dense settings.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 101985"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spatial modeling of geological and atmospheric hazards for volcanic eruption mitigation at Mount Lewotobi Laki-laki","authors":"Devina Fauziah,&nbsp;Syaefunnisa Syaefunnisa,&nbsp;Fikri Algifahri,&nbsp;Lia Berliana,&nbsp;Muhammad Taufikurohman,&nbsp;Alfarizy Fajril Maulad,&nbsp;Tania Septi Anggraini,&nbsp;Silmi Afina Aliyan","doi":"10.1016/j.rsase.2026.101999","DOIUrl":"10.1016/j.rsase.2026.101999","url":null,"abstract":"<div><div>Mount Lewotobi Laki-laki is one of Indonesia's active volcanoes, exhibiting significant activity during October–December 2024. This research aims to identify priority areas for mitigation based on geological and atmospheric hazards associated with this eruption event. First, this research identifies geological (primary) hazards by integrating a Digital Elevation Model (DEM), volcanic disaster prone area data, river data, and distance from the crater, all of which are analyzed using Geographic Information System (GIS). Second, this study identifies the atmospheric (secondary) hazard, utilizing the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) Dispersion model and the Differential Optical Absorption Spectroscopy (DOAS) technique. Third, this research identifies the total hazard by integrating primary and secondary hazards. Fourth, this research models the prioritization area for eruption mitigation by integrating total hazard and exposure data. From modeling, this research obtains some results (1) primary hazards are concentrated around the crater and along river channels, with Wulanggitang District as the most vulnerable area; (2) secondary hazards show that Sulfur Dioxide (SO₂) gas and volcanic ash spread widely to the west, reaching Lombok and Bali during the peak eruption; (3) total hazards indicate that most of the area falls into the high category, but very high hazard zones remain significant around the crater (4) priority mitigation areas are Wulanggitang and Ile Bura Districts which require early evacuation. The use of GIS and remote sensing substantially enhances the prioritization of disaster mitigation efforts in areas affected by volcanic eruptions. The results are specific to the October–December 2024 eruption and should be interpreted as a screening-level prioritization. Applying this framework to other eruptions requires updated hazard inputs and recalibration.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 101999"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimizing floriculture mapping with deep learning using high-resolution multispectral and hyperspectral imagery","authors":"Subhajit Patra,&nbsp;Dipanwita Dutta,&nbsp;Shrabani Kar,&nbsp;Debarati Bera","doi":"10.1016/j.rsase.2026.101968","DOIUrl":"10.1016/j.rsase.2026.101968","url":null,"abstract":"<div><div>Accurate mapping of floriculture is essential for efficient agricultural management and sustainable practices in the flower industry. However, conventional remote sensing techniques are not sufficient for mapping floriculture in diverse, heterogenous landscape. This study explores the potentiality of machine learning and deep learning models for identification of floriculture areas, specifically, rose farms integrating with high-resolution multispectral and hyperspectral datasets. The PlanetScope SuperDove and PRISMA data were employed independently and as fused datasets, incorporating multiple vegetation indices, including NDVI, SAVI, OSAVI, and NDWI, to enhance model accuracy. Spectral signatures of rose and marigold flowers along with GPS based sample points were collected during the field survey. The classification models evaluated in this study included Random Forest (RF), Support Vector Machine (SVM), Convolutional Neural Network (CNN), and Deep Neural Network (DNN), each assessed for their ability to differentiate various features within complex, heterogeneous floriculture farms. This study revealed that the fusion of high spectral and spatial resolution data significantly improves classification accuracy. Deep learning model, especially when integrated with fused datasets, achieved notable performance metrics, an overall accuracy of 83% and a Kappa coefficient of 78% surpassing traditional machine learning approaches. This model demonstrated a higher recall of 0.86 and an F1-score of 0.83, indicating its superior ability to accurately distinguish rose farms from other crops with subtle spectral differences. Integration of high resolution multi-spectral data with the PRISMA hyperspectral image enhanced the potentiality of GeoAI based models. This research would be helpful for policy intervention, resource allocation and environmental management strategies, facilitating more efficient agricultural practices and thereby, promoting sustainable practices.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 101968"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613420","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mapping and spatiotemporal dynamics of aquatic vegetation in the seasonal sub-lakes of Poyang Lake using phenological features","authors":"Wei Liang ,&nbsp;Runyuan Kuang ,&nbsp;Jiabao Tan ,&nbsp;Jiaxin Qiao","doi":"10.1016/j.rsase.2026.101995","DOIUrl":"10.1016/j.rsase.2026.101995","url":null,"abstract":"<div><div>Aquatic vegetation is a vital component of seasonal lake ecosystems. Its accurate extraction under the changing environment of water and land is of significant theoretical and practical importance for diagnosing ecosystem health, analyzing its response mechanisms to extreme climate events, and ultimately serving regional adaptive ecological management. This study focuses on the seasonal sub-lakes within Poyang Lake. It develops a hybrid method based on Sentinel-2 optical imagery that combines rule-based constraints with machine learning to achieve precise extraction and spatiotemporal analysis of aquatic vegetation. The results indicate that: (1) The extraction method demonstrated high accuracy and stability, with Overall Accuracy (OA) values ranging from 0.943 to 0.987 and Kappa coefficients from 0.820 to 0.940 for the period 2019-2024. (2) Time series analysis revealed a significant declining trend in the area of aquatic vegetation in Poyang Lake from 2019 to 2024. A strong correlation was identified between the extreme flood event in 2020 and the sharp decline in aquatic vegetation area in 2021, whereas the 2022 drought event had a relatively minor impact. Spatially, aquatic vegetation is mainly distributed in the Nanji Wetland National Nature Reserve and the Poyang Lake National Nature Reserve within the study area. (3) The growth of aquatic vegetation was found to be significantly and positively correlated with moderate water depth, specific water-level fluctuation patterns, and ambient temperature. The method proposed in this study effectively overcomes the challenges brought by the complex environment of seasonal lakes, and its research results will provide important scientific support for the formulation of ecological restoration and adaptive management strategies in the region.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"42 ","pages":"Article 101995"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147613424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}