Remote Sensing Applications-Society and Environment最新文献_第3页

Quantifying human-induced impacts on forest phenology using multi-source remote sensing data: A case study in the Gudao Oilfield, China 利用多源遥感数据量化人类活动对森林物候的影响——以孤岛油田为例

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-01-08 DOI: 10.1016/j.rsase.2026.101879

Han Yang , Hong Wang , Nobuaki Tanaka

Vegetation phenology reflects the seasonal dynamics of ecosystems. It responds to global climate change and is significantly influenced by local human activities. However, the spatial dimension of human-induced phenological impacts remains unclear. This study employed PlanetScope (PS, 3 m), Sentinel-2 (S2, 10 m), and Harmonized Landsat Sentinel-2 (HLS, 30 m) imagery to quantify the impacts of oil extraction and road-related activities on shelterbelt phenology in the Gudao Oilfield. Phenological changes within distance-based buffers around the disturbance sources were modelled using an exponential decay function to derive cumulative impact curves. Based on the Pareto principle, the distance at which the cumulative impact reached 80 % and the corresponding phenological change were used to characterize these human-induced impacts. Results showed that human activities advanced the start (SOS) and delayed the end (EOS) of the growing season compared with reference areas (>300 m from the road and >200 m from all oil wells). Estimated influence distances from PS and S2 imagery were 37.57–51.00 m for road-related activities and 38.93–43.43 m for oil extraction, comparable to the observed spatial extent of forest structural changes, with corresponding phenological changes of 2.40–3.91 and 4.50–6.65 days, respectively. Scale effects introduced uncertainty in quantifying human-induced impacts. At the coarser 30 m resolution (HLS imagery), influence distances were overestimated (>83.07 m) and phenological changes were underestimated (<1.94 days). This study provides a methodological framework for quantifying human-induced impacts on vegetation phenology and offers new insights into scale effects in ecological monitoring.

植被物候反映了生态系统的季节动态。它对全球气候变化作出反应，并受到当地人类活动的重大影响。然而，人为物候影响的空间维度尚不清楚。利用PlanetScope （PS, 3 m）、Sentinel-2 （S2, 10 m）和Harmonized Landsat Sentinel-2 （HLS, 30 m）影像，量化了采油和道路相关活动对孤岛油田林带物候的影响。在干扰源周围基于距离的缓冲区内的物候变化使用指数衰减函数建模，以得出累积影响曲线。基于Pareto原理，利用累积影响达到80%的距离和相应的物候变化特征来表征这些人为影响。结果表明：与参考区（距公路300 m，距所有油井200 m）相比，人类活动使生长季的开始（SOS）提前，结束（EOS）推迟；PS和S2影像对道路活动和采油活动的影响距离分别为37.57 ~ 51.00 m和38.93 ~ 43.43 m，与观测到的森林结构变化空间范围相当，对应的物候变化分别为2.40 ~ 3.91天和4.50 ~ 6.65天。尺度效应在量化人为影响时引入了不确定性。在较粗的30米分辨率（HLS图像）下，影响距离被高估（83.07米），物候变化被低估（1.94天）。该研究为量化人类活动对植被物候的影响提供了一个方法框架，并为生态监测中的尺度效应提供了新的见解。

{"title":"Quantifying human-induced impacts on forest phenology using multi-source remote sensing data: A case study in the Gudao Oilfield, China","authors":"Han Yang , Hong Wang , Nobuaki Tanaka","doi":"10.1016/j.rsase.2026.101879","DOIUrl":"10.1016/j.rsase.2026.101879","url":null,"abstract":"<div><div>Vegetation phenology reflects the seasonal dynamics of ecosystems. It responds to global climate change and is significantly influenced by local human activities. However, the spatial dimension of human-induced phenological impacts remains unclear. This study employed PlanetScope (PS, 3 m), Sentinel-2 (S2, 10 m), and Harmonized Landsat Sentinel-2 (HLS, 30 m) imagery to quantify the impacts of oil extraction and road-related activities on shelterbelt phenology in the Gudao Oilfield. Phenological changes within distance-based buffers around the disturbance sources were modelled using an exponential decay function to derive cumulative impact curves. Based on the Pareto principle, the distance at which the cumulative impact reached 80 % and the corresponding phenological change were used to characterize these human-induced impacts. Results showed that human activities advanced the start (SOS) and delayed the end (EOS) of the growing season compared with reference areas (>300 m from the road and >200 m from all oil wells). Estimated influence distances from PS and S2 imagery were 37.57–51.00 m for road-related activities and 38.93–43.43 m for oil extraction, comparable to the observed spatial extent of forest structural changes, with corresponding phenological changes of 2.40–3.91 and 4.50–6.65 days, respectively. Scale effects introduced uncertainty in quantifying human-induced impacts. At the coarser 30 m resolution (HLS imagery), influence distances were overestimated (>83.07 m) and phenological changes were underestimated (<1.94 days). This study provides a methodological framework for quantifying human-induced impacts on vegetation phenology and offers new insights into scale effects in ecological monitoring.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101879"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977176","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}

引用次数: 0

Unraveling mangrove degradation in Jardines de la Reina National Park, Cuba: Integration of Landsat-8, machine learning and environmental factors 解开古巴怡和雷纳国家公园红树林退化：Landsat-8、机器学习和环境因素的整合

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-01-09 DOI: 10.1016/j.rsase.2025.101861

Alexey Valero-Jorge , Roberto González De-Zayas , Angel Luis Becerra González , Felipe Matos-Pupo , Dian Nuraini Melati , Eduardo González-Ferreiro

Mangrove ecosystems are vital for coastal resilience, biodiversity, and climate regulation. This study assessed the spatial and temporal dynamics of mangrove cover in Jardines de la Reina National Park (JRNP), Cuba, between 2014 and 2024, using Landsat-8 imagery and five machine learning classifiers. Random Forest (RF) achieved the highest accuracy (97.60 %), with rigorous uncertainty propagation via Monte Carlo simulation, setting a new benchmark for mangrove mapping in the data-poor insular Caribbean. This method was selected for generating annual maps. Results revealed a loss of over 1500 ha of mangrove forest—an 18.65 % reduction—primarily in the western sector, especially Bretón and Alcatraz keys. NDVI trend analysis confirmed significant degradation in these areas, while central keys remained more stable. Environmental factor analysis identified mean sea level (MSL) as the dominant driver of mangrove loss, followed by annual precipitation. Limited freshwater and sediment input, exacerbated by damming and droughts, likely impaired mangrove resilience. Patchy dieback patterns were observed, with localized mortality within otherwise healthy stands. Herbivory by hutia (Capromys pilorides) may contribute to stress, but recent data are lacking. Although JRNP is a protected area, the dominance of external environmental drivers—particularly sea-level rise and reduced precipitation—poses challenges that may exceed current local conservation management capabilities. The study highlights the need for integrated field and remote sensing approaches to monitor ecosystem health. Future research should focus on sediment accretion, primary productivity, herbivory impacts, and hydrological connectivity. This framework offers a model for holistic mangrove in marine protected areas across the Caribbean and supports adaptive management strategies to address the challenges posed by climate change.

红树林生态系统对沿海恢复力、生物多样性和气候调节至关重要。本研究利用Landsat-8卫星图像和五种机器学习分类器，评估了2014年至2024年间古巴雷纳花园国家公园（JRNP）红树林覆盖的时空动态。随机森林（RF）通过蒙特卡罗模拟获得了最高的精度（97.60%），具有严格的不确定性传播，为数据贫乏的加勒比岛屿红树林制图设定了新的基准。选择这种方法生成年度地图。结果显示，损失了超过1500公顷的红树林，减少了18.65%，主要是在西部地区，特别是Bretón和阿尔卡特拉斯群岛。NDVI趋势分析证实了这些区域的显著退化，而中央键保持较稳定。环境因子分析表明，平均海平面是红树林损失的主要驱动因素，其次是年降水量。有限的淡水和沉积物输入，再加上筑坝和干旱，可能损害了红树林的恢复能力。观察到斑驳的枯死模式，在其他健康的林分中有局部死亡。竹属植物（Capromys pilorides）的食草性可能导致压力，但缺乏最近的数据。虽然JRNP是一个保护区，但外部环境驱动因素（特别是海平面上升和降水减少）的主导地位带来的挑战可能超出当前当地的保护管理能力。该研究强调需要采用综合的野外和遥感方法来监测生态系统健康。未来的研究应集中在泥沙增积、初级生产力、草食影响和水文连通性等方面。该框架为整个加勒比海洋保护区的整体红树林提供了一个模式，并支持适应性管理战略，以应对气候变化带来的挑战。

{"title":"Unraveling mangrove degradation in Jardines de la Reina National Park, Cuba: Integration of Landsat-8, machine learning and environmental factors","authors":"Alexey Valero-Jorge , Roberto González De-Zayas , Angel Luis Becerra González , Felipe Matos-Pupo , Dian Nuraini Melati , Eduardo González-Ferreiro","doi":"10.1016/j.rsase.2025.101861","DOIUrl":"10.1016/j.rsase.2025.101861","url":null,"abstract":"<div><div>Mangrove ecosystems are vital for coastal resilience, biodiversity, and climate regulation. This study assessed the spatial and temporal dynamics of mangrove cover in Jardines de la Reina National Park (JRNP), Cuba, between 2014 and 2024, using Landsat-8 imagery and five machine learning classifiers. Random Forest (RF) achieved the highest accuracy (97.60 %), with rigorous uncertainty propagation via Monte Carlo simulation, setting a new benchmark for mangrove mapping in the data-poor insular Caribbean. This method was selected for generating annual maps. Results revealed a loss of over 1500 ha of mangrove forest—an 18.65 % reduction—primarily in the western sector, especially Bretón and Alcatraz keys. NDVI trend analysis confirmed significant degradation in these areas, while central keys remained more stable. Environmental factor analysis identified mean sea level (MSL) as the dominant driver of mangrove loss, followed by annual precipitation. Limited freshwater and sediment input, exacerbated by damming and droughts, likely impaired mangrove resilience. Patchy dieback patterns were observed, with localized mortality within otherwise healthy stands. Herbivory by hutia (<em>Capromys pilorides</em>) may contribute to stress, but recent data are lacking. Although JRNP is a protected area, the dominance of external environmental drivers—particularly sea-level rise and reduced precipitation—poses challenges that may exceed current local conservation management capabilities. The study highlights the need for integrated field and remote sensing approaches to monitor ecosystem health. Future research should focus on sediment accretion, primary productivity, herbivory impacts, and hydrological connectivity. This framework offers a model for holistic mangrove in marine protected areas across the Caribbean and supports adaptive management strategies to address the challenges posed by climate change.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101861"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977873","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}

引用次数: 0

Influence of atmospheric boundary-layer dynamics on air quality of the middle- and high-density urban areas of Colombia 大气边界层动力学对哥伦比亚中部和高密度城区空气质量的影响

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-01-07 DOI: 10.1016/j.rsase.2026.101874

Luis M. Hernández Beleño , Gregori de Arruda Moreira , Eliana Vergara-Vásquez , Yiniva Camargo Caicedo , David J. O'Connor , Andrés M. Vélez-Pereira

The interplay between emissions and atmospheric boundary-layer dynamics shapes urban air quality (AQ) in Colombia's complex topography. This study assesses the influence of the atmospheric boundary layer on AQ across contrasting physiographic regions. The ERA5 reanalysis dataset was used to obtain hourly ABLH and VC estimates for the period 2020–2024, while COSMIC-2 profiles were used to derive Temperature Elevation Profile (TEP) variables, including inversion-base height and thermal gradients. Urban AQ data from 78 monitoring stations were obtained from SISAIRE, focusing on PM₁₀, PM_2.5, and O₃. The analysis combines exceedance rates (98th-percentile thresholds), diurnal and seasonal cycles, nonparametric correlations, and Gaussian linear models stratified by stable/unstable ABL conditions and dry/wet seasons. Our results show frequent exceedances in Antioquia and Bogotá, where PM_2.5 daily exceedance medians reach 1.11 % and 0.87 %, respectively. Norte de Santander exhibits the highest PM_2.5 median exceedance rate (7.18 %), while departments such as Cesar and Magdalena show low-to-moderate levels. O₃ responses are strongly modulated by thermal structure, with direct associations between ABLH, inversion strength, and O₃ peaks, particularly in high-elevation terrains. Physiography and circulation patterns explain regional contrasts, with stagnation-prone basins showing stronger pollution accumulation. We conclude that ventilation conditions strongly influence particulate pollution, whereas peak O₃ is governed primarily by precursor emissions and temperature-driven photochemistry. These findings highlight the need for meteorology-aware AQ management strategies, especially in densely populated Andean basins.

排放和大气边界层动力学之间的相互作用决定了哥伦比亚复杂地形下的城市空气质量。本研究评估了不同地理区域大气边界层对空气质量的影响。ERA5再分析数据集用于获得2020-2024年每小时ABLH和VC估计，而COSMIC-2剖面用于获得温度高程剖面（TEP）变量，包括反演基高和热梯度。来自SISAIRE的78个监测站的城市空气质量数据，重点关注PM10、PM2.5和O3。该分析结合了超过率（第98百分位阈值）、日和季节周期、非参数相关性以及由稳定/不稳定ABL条件和干/湿季节分层的高斯线性模型。我们的研究结果显示，安蒂奥基亚和波哥大的PM2.5日超标中位数分别达到1.11%和0.87%。北桑坦德的PM2.5中位数超标率最高（7.18%），而凯撒和马格达莱纳等省的PM2.5中位数超标率为中低水平。O3响应受到热结构的强烈调节，在ABLH、逆温强度和O3峰值之间存在直接关联，特别是在高海拔地区。地形和环流模式解释了区域差异，容易停滞的盆地表现出更强的污染积累。我们得出结论，通风条件强烈影响颗粒污染，而O3峰值主要由前体排放和温度驱动的光化学控制。这些发现突出表明需要有气象意识的空气质量管理策略，特别是在人口稠密的安第斯盆地。

{"title":"Influence of atmospheric boundary-layer dynamics on air quality of the middle- and high-density urban areas of Colombia","authors":"Luis M. Hernández Beleño , Gregori de Arruda Moreira , Eliana Vergara-Vásquez , Yiniva Camargo Caicedo , David J. O'Connor , Andrés M. Vélez-Pereira","doi":"10.1016/j.rsase.2026.101874","DOIUrl":"10.1016/j.rsase.2026.101874","url":null,"abstract":"<div><div>The interplay between emissions and atmospheric boundary-layer dynamics shapes urban air quality (AQ) in Colombia's complex topography. This study assesses the influence of the atmospheric boundary layer on AQ across contrasting physiographic regions. The ERA5 reanalysis dataset was used to obtain hourly ABLH and VC estimates for the period 2020–2024, while COSMIC-2 profiles were used to derive Temperature Elevation Profile (TEP) variables, including inversion-base height and thermal gradients. Urban AQ data from 78 monitoring stations were obtained from SISAIRE, focusing on PM<sub>10</sub>, PM<sub>2.5</sub>, and O<sub>3</sub>. The analysis combines exceedance rates (98th-percentile thresholds), diurnal and seasonal cycles, nonparametric correlations, and Gaussian linear models stratified by stable/unstable ABL conditions and dry/wet seasons. Our results show frequent exceedances in Antioquia and Bogotá, where PM<sub>2.5</sub> daily exceedance medians reach 1.11 % and 0.87 %, respectively. Norte de Santander exhibits the highest PM<sub>2.5</sub> median exceedance rate (7.18 %), while departments such as Cesar and Magdalena show low-to-moderate levels. O<sub>3</sub> responses are strongly modulated by thermal structure, with direct associations between ABLH, inversion strength, and O<sub>3</sub> peaks, particularly in high-elevation terrains. Physiography and circulation patterns explain regional contrasts, with stagnation-prone basins showing stronger pollution accumulation. We conclude that ventilation conditions strongly influence particulate pollution, whereas peak O<sub>3</sub> is governed primarily by precursor emissions and temperature-driven photochemistry. These findings highlight the need for meteorology-aware AQ management strategies, especially in densely populated Andean basins.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101874"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926096","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}

引用次数: 0

Spatio-temporal analysis of land use and land cover (LULC) dynamics: Trends, drivers and implications in semi-arid Vanivilasa sagara reservoir catchment, India 印度半干旱Vanivilasa sagara水库流域土地利用和土地覆盖动态的时空分析：趋势、驱动因素及其影响

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2025-12-09 DOI: 10.1016/j.rsase.2025.101816

Lokanath S, Govindaraju, Rakesh C J, Kishor Kumar A

Land use and land cover (LULC) changes induce unrivalled environmental changes at various spatial and temporal scales, affecting land surface processes in present catchment and it has experienced detrimental LULC changes over the past three decades, characterized by mismanagement of land and water resources. Hence, the current investigation aimed to analyze the spatial–temporal dynamics of LULC changes from 1990 to 2023 and to predict future trend scenario (1990–2053) for the Vanivilasa Sagara reservoir catchment, Karnataka state in India. The analysis was employed using integrated remote sensing and multitemporal geospatial data by utilizing Landsat – 5, Landsat – 7, IRS LISS – III, and Sentinel – 2A data, with aligning spatial resolutions, uncertainties, along with cumulative statistical model, and ground survey. First, a hybrid–image classification technique was employed to create LULC maps spanning 33 years (1990–2023) across seven reference periods: 1990, 1995, 2000, 2007, 2015, 2020, and 2023, and validated by accuracy assessment using meticulous field data. Subsequently, the magnitude, extent, trajectories, change rate, overall gains, losses, and net LULC changes were derived using statistical evaluation. The concerted results have indicated significant LULC alterations, including transforming forests, scrublands, and grasslands into agricultural and built-up regions such as human settlements, industries, and quarries/mining; cropping pattern conversions; marked rise in wasteland ecosystems such as erosional gullied land, salt-affected land, marshes, and land without scrubs. Moreover, unforeseen thrive in surface water bodies has been found. Key driving parameters and implications of LULC dynamics that hamper the environment of the study area were comprehensively studied and deciphered. Further, future LULC trend analysis obtained by Visual Basics Analysis (VBA) module, serves as an early warning system for resource degradation in catchments amidst climate change is addressed by this study, which contributes to regional planning between researchers, farmers, and government authorities at every level to achieve harmony.

土地利用和土地覆盖（LULC）的变化在不同的时空尺度上引起了无与伦比的环境变化，影响了当前流域的陆面过程，并且在过去的30年中经历了有害的LULC变化，其特征是土地和水资源管理不善。因此，本研究旨在分析印度卡纳塔克邦Vanivilasa Sagara水库集水区1990 - 2023年LULC变化的时空动态，并预测其未来趋势情景（1990 - 2053年）。利用Landsat - 5、Landsat - 7、IRS LISS - III和Sentinel - 2A数据，结合空间分辨率、不确定性、累积统计模型和地面调查数据，利用遥感和多时相地理空间数据进行分析。首先，采用混合图像分类技术，在1990年、1995年、2000年、2007年、2015年、2020年和2023年7个参考时期，创建了33年（1990 - 2023年）的LULC地图，并利用细致的实地数据进行了精度评估。随后，通过统计评估得出了幅度、范围、轨迹、变化率、总体收益、损失和净LULC变化。协调一致的结果表明了重大的土地利用变化，包括将森林、灌木丛和草原转变为农业和建筑区域，如人类住区、工业和采石场/采矿；种植模式转换；荒地生态系统显著增加，如侵蚀沟壑地、盐渍地、沼泽和没有灌木的土地。此外，在地表水体中发现了意想不到的繁盛。全面研究和解读了影响研究区生态环境的LULC动态的关键驱动参数和影响。此外，本研究利用VBA （Visual Basics analysis）模块获得的未来LULC趋势分析，作为气候变化背景下流域资源退化的预警系统，有助于研究人员、农民和各级政府部门之间的区域规划，实现和谐。

{"title":"Spatio-temporal analysis of land use and land cover (LULC) dynamics: Trends, drivers and implications in semi-arid Vanivilasa sagara reservoir catchment, India","authors":"Lokanath S, Govindaraju, Rakesh C J, Kishor Kumar A","doi":"10.1016/j.rsase.2025.101816","DOIUrl":"10.1016/j.rsase.2025.101816","url":null,"abstract":"<div><div>Land use and land cover (LULC) changes induce unrivalled environmental changes at various spatial and temporal scales, affecting land surface processes in present catchment and it has experienced detrimental LULC changes over the past three decades, characterized by mismanagement of land and water resources. Hence, the current investigation aimed to analyze the spatial–temporal dynamics of LULC changes from 1990 to 2023 and to predict future trend scenario (1990–2053) for the Vanivilasa Sagara reservoir catchment, Karnataka state in India. The analysis was employed using integrated remote sensing and multitemporal geospatial data by utilizing Landsat – 5, Landsat – 7, IRS LISS – III, and Sentinel – 2A data, with aligning spatial resolutions, uncertainties, along with cumulative statistical model, and ground survey. First, a hybrid–image classification technique was employed to create LULC maps spanning 33 years (1990–2023) across seven reference periods: 1990, 1995, 2000, 2007, 2015, 2020, and 2023, and validated by accuracy assessment using meticulous field data. Subsequently, the magnitude, extent, trajectories, change rate, overall gains, losses, and net LULC changes were derived using statistical evaluation. The concerted results have indicated significant LULC alterations, including transforming forests, scrublands, and grasslands into agricultural and built-up regions such as human settlements, industries, and quarries/mining; cropping pattern conversions; marked rise in wasteland ecosystems such as erosional gullied land, salt-affected land, marshes, and land without scrubs. Moreover, unforeseen thrive in surface water bodies has been found. Key driving parameters and implications of LULC dynamics that hamper the environment of the study area were comprehensively studied and deciphered. Further, future LULC trend analysis obtained by Visual Basics Analysis (VBA) module, serves as an early warning system for resource degradation in catchments amidst climate change is addressed by this study, which contributes to regional planning between researchers, farmers, and government authorities at every level to achieve harmony.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101816"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145790943","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}

引用次数: 0

Driving mechanisms of nitrogen and phosphorus dynamics in the Daitou River basin: A multi-level perspective from Sentinel-2 imagery 大头河流域氮磷动态的驱动机制：基于Sentinel-2遥感影像的多层次视角

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2025-12-22 DOI: 10.1016/j.rsase.2025.101848

Yuanmao Zheng , Chenyan Wei , Lingluo Chen , Haiyan Fu , Haoxi Lin

It was found by the Central Environmental Protection Inspection Team of China that the Daitou River Basin in Tong' an District, Xiamen City, had prominent issues concerning its black and odorous water bodies. To explore the relationships between natural factors, socioeconomic development factors, and land use pattern factors in the Daitou River Basin and the concentrations of nitrogen (TN) and phosphorus (TP) in the water bodies, this study integrated monitoring data on TN and TP concentrations with statistical data to construct a model of the impact mechanisms of TN and TP concentrations. The study revealed the factors influencing TN and TP concentrations and their impact mechanisms and proposed corresponding treatment strategies and recommendations. Based on the geographical detector model, TN and TP concentrations were found to be significantly influenced by precipitation, runoff, population size, GDP, and the primary industry value. Among these drivers, the most pronounced effect on TN concentration was exerted by precipitation, with an explanatory power of 0.9376, whereas TP concentration was most strongly affected by GDP, reaching an explanatory power of 0.9777. According to the correlation-based analytical model, precipitation, population size, and the primary industry value were considered as the dominant factors governing the spatiotemporal distinction of TN and TP concentrations. Specifically, the explanatory power for TN concentration was observed to range from 0.618 to 0.878 for precipitation, 0.765 to 0.873 for the primary industry value, and 0.642 to 0.785 for population size; for TP concentration, these values were 0.692–0.876, 0.642–0.874, and 0.551–0.869, respectively. The research results of this study can provide more scientific and solid theoretical basis for formulating water ecological treatment measures.

近日，中央环境保护督察组发现，厦门市通安区大头河流域水体黑臭问题突出。为探索岱头河流域自然因素、社会经济发展因素和土地利用模式因素与水体中氮、磷浓度的关系，本研究将TN、TP浓度监测数据与统计数据相结合，构建了TN、TP浓度影响机制模型。本研究揭示了影响总氮和总磷浓度的因素及其影响机制，并提出了相应的处理策略和建议。基于地理探测器模型，发现全氮和总磷浓度受降水、径流、人口规模、GDP和第一产业价值的显著影响。其中，降水对全氮浓度的影响最为显著，解释能力为0.9376，而全磷浓度受GDP的影响最为强烈，解释能力为0.9777。根据相关性分析模型，降水量、人口规模和第一产业价值是控制全氮和总磷浓度时空差异的主导因素。其中，降水量对TN浓度的解释能力为0.618 ~ 0.878，第一产业值对TN浓度的解释能力为0.765 ~ 0.873，人口规模对TN浓度的解释能力为0.642 ~ 0.785；TP浓度分别为0.692 ~ 0.876、0.642 ~ 0.874和0.551 ~ 0.869。本研究成果可为制定水生态治理措施提供更为科学、坚实的理论依据。

{"title":"Driving mechanisms of nitrogen and phosphorus dynamics in the Daitou River basin: A multi-level perspective from Sentinel-2 imagery","authors":"Yuanmao Zheng , Chenyan Wei , Lingluo Chen , Haiyan Fu , Haoxi Lin","doi":"10.1016/j.rsase.2025.101848","DOIUrl":"10.1016/j.rsase.2025.101848","url":null,"abstract":"<div><div>It was found by the Central Environmental Protection Inspection Team of China that the Daitou River Basin in Tong' an District, Xiamen City, had prominent issues concerning its black and odorous water bodies. To explore the relationships between natural factors, socioeconomic development factors, and land use pattern factors in the Daitou River Basin and the concentrations of nitrogen (TN) and phosphorus (TP) in the water bodies, this study integrated monitoring data on TN and TP concentrations with statistical data to construct a model of the impact mechanisms of TN and TP concentrations. The study revealed the factors influencing TN and TP concentrations and their impact mechanisms and proposed corresponding treatment strategies and recommendations. Based on the geographical detector model, TN and TP concentrations were found to be significantly influenced by precipitation, runoff, population size, GDP, and the primary industry value. Among these drivers, the most pronounced effect on TN concentration was exerted by precipitation, with an explanatory power of 0.9376, whereas TP concentration was most strongly affected by GDP, reaching an explanatory power of 0.9777. According to the correlation-based analytical model, precipitation, population size, and the primary industry value were considered as the dominant factors governing the spatiotemporal distinction of TN and TP concentrations. Specifically, the explanatory power for TN concentration was observed to range from 0.618 to 0.878 for precipitation, 0.765 to 0.873 for the primary industry value, and 0.642 to 0.785 for population size; for TP concentration, these values were 0.692–0.876, 0.642–0.874, and 0.551–0.869, respectively. The research results of this study can provide more scientific and solid theoretical basis for formulating water ecological treatment measures.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101848"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926035","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}

引用次数: 0

Comparative analysis of PS-InSAR and DS-InSAR for deformation monitoring of transportation infrastructure in Greater Bay Area, China PS-InSAR与DS-InSAR在大湾区交通基础设施变形监测中的对比分析

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-01-05 DOI: 10.1016/j.rsase.2026.101872

Songbo Wu , Bochen Zhang , Yan Li , Siting Xiong , Xiaoli Ding

Transportation networks are vital for our economy, such as the Guangdong-Hong Kong-Macao Greater Bay Area (GBA). However, their large scale makes them easily susceptible to the ground deformation. Synthetic Aperture Radar Interferometry (InSAR), as an efficient geodetic technique, has been widely used for regional surface deformation monitoring. However, choosing the best InSAR method for effective monitoring of diverse transportation infrastructure remains a key challenge. This study aims to address this issue in GBA region by conducting a systematic multi-sensor comparison of two widely used InSAR methods, i.e., Permanent Scatterer (PS-InSAR) and Distributed Scatterer (DS-InSAR). Based on three SAR data from Sentinel-1A, COSMO-SkyMed, and PALSAR-2, we conducted the ground deformation monitoring and statistical analysis for various infrastructure types (including high-speed railways, bridges, coastal highways, and airports) within the GBA. The spatial distribution, density, and coverage of the measurements, were evaluated. The experimental results were validated against GNSS benchmark data, confirming the reliability of the measurements. We quantitatively demonstrate that in urban areas of GBA, the suitability of a given technique depends primarily on the surface characteristics of the target and its surrounding environment. DS-InSAR performs better in low coherence region e.g., the construction zones and low-reflectivity pavements, achieving denser point than PS-InSAR. But it requires significantly more computation. In contrast, PS-InSAR effectively detects deformation hotspots and provides high-accuracy for stable linear structures such as cross-sea bridges. We further quantified the influence of key monitoring parameters, including observation period, sensor wavelength, and ground vegetation characteristics and compared their roles in monitoring the transportation infrastructure network. The study results provide a comprehensive evaluation of the monitoring effectiveness and efficiency of those two methods, which supports the selection of an optimal InSAR approach for future applications in the GBA.

交通网络对香港经济至关重要，例如粤港澳大湾区。然而，它们的规模大，容易受到地面变形的影响。合成孔径雷达干涉测量技术作为一种有效的大地测量技术，在区域地表变形监测中得到了广泛的应用。然而，选择最佳的InSAR方法来有效监测各种交通基础设施仍然是一个关键的挑战。本研究旨在通过对两种广泛使用的InSAR方法，即永久散射体（PS-InSAR）和分布式散射体（DS-InSAR）进行系统的多传感器比较，解决大湾区地区的这一问题。基于Sentinel-1A、COSMO-SkyMed和PALSAR-2卫星SAR数据，对大湾区内各类基础设施类型（包括高速铁路、桥梁、沿海公路和机场）进行了地面变形监测和统计分析。评估了测量的空间分布、密度和覆盖范围。实验结果与GNSS基准数据进行了验证，验证了测量结果的可靠性。我们定量地证明，在大湾区的城市地区，给定技术的适用性主要取决于目标及其周围环境的表面特征。DS-InSAR在低相干区域（如施工区域和低反射率路面）表现更好，比PS-InSAR获得更密集的点。但它需要更多的计算。相比之下，PS-InSAR可以有效地检测变形热点，并为跨海桥梁等稳定的线性结构提供高精度。我们进一步量化了关键监测参数的影响，包括观测周期、传感器波长和地面植被特征，并比较了它们在监测交通基础设施网络中的作用。研究结果对这两种方法的监测效果和效率进行了综合评价，为未来在大湾区的应用选择最佳的InSAR方法提供了支持。

{"title":"Comparative analysis of PS-InSAR and DS-InSAR for deformation monitoring of transportation infrastructure in Greater Bay Area, China","authors":"Songbo Wu , Bochen Zhang , Yan Li , Siting Xiong , Xiaoli Ding","doi":"10.1016/j.rsase.2026.101872","DOIUrl":"10.1016/j.rsase.2026.101872","url":null,"abstract":"<div><div>Transportation networks are vital for our economy, such as the Guangdong-Hong Kong-Macao Greater Bay Area (GBA). However, their large scale makes them easily susceptible to the ground deformation. Synthetic Aperture Radar Interferometry (InSAR), as an efficient geodetic technique, has been widely used for regional surface deformation monitoring. However, choosing the best InSAR method for effective monitoring of diverse transportation infrastructure remains a key challenge. This study aims to address this issue in GBA region by conducting a systematic multi-sensor comparison of two widely used InSAR methods, i.e., Permanent Scatterer (PS-InSAR) and Distributed Scatterer (DS-InSAR). Based on three SAR data from Sentinel-1A, COSMO-SkyMed, and PALSAR-2, we conducted the ground deformation monitoring and statistical analysis for various infrastructure types (including high-speed railways, bridges, coastal highways, and airports) within the GBA. The spatial distribution, density, and coverage of the measurements, were evaluated. The experimental results were validated against GNSS benchmark data, confirming the reliability of the measurements. We quantitatively demonstrate that in urban areas of GBA, the suitability of a given technique depends primarily on the surface characteristics of the target and its surrounding environment. DS-InSAR performs better in low coherence region e.g., the construction zones and low-reflectivity pavements, achieving denser point than PS-InSAR. But it requires significantly more computation. In contrast, PS-InSAR effectively detects deformation hotspots and provides high-accuracy for stable linear structures such as cross-sea bridges. We further quantified the influence of key monitoring parameters, including observation period, sensor wavelength, and ground vegetation characteristics and compared their roles in monitoring the transportation infrastructure network. The study results provide a comprehensive evaluation of the monitoring effectiveness and efficiency of those two methods, which supports the selection of an optimal InSAR approach for future applications in the GBA.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101872"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926036","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}

引用次数: 0

Urban local climate zone classification through deep learning using spatio-temporal thermal imagery 基于时空热像的深度学习城市局地气候带分类

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-01-22 DOI: 10.1016/j.rsase.2026.101889

Michaja van Capel , Azarakhsh Rafiee , Roderik Lindenbergh

Rapid urbanization challenges urban micro-climates, strains resources and affects public health. Understanding micro-climate dynamics is key to effective mitigation and sustainable development. Local Climate Zone (LCZ) classification supports climate-resilient planning but is complicated by the diversity and complexity of diverse urban landscapes and the coexistence of varying land uses and materials within small areas. While LCZ classification typically uses multispectral imagery, LiDAR, and land-use data, these sources often miss temporal thermal dynamic patterns. Thermal satellite imagery improves LCZ classification by distinguishing zones with similar structures but differing thermal behavior. This research proposes using deep learning-based multitemporal semantic segmentation to classify urban LCZs based solely on temporal thermal patterns from ECOSTRESS satellite imagery. The methodology is applied in a in a case study around the near coastal cities of Rotterdam and The Hague in The Netherlands and demonstrates how spatial and temporal factors (both diurnal and seasonal) influence the performance of the semantic segmentation model on different LCZ classes. The study shows that a U-Net architecture applied on spatio-temporal thermal imagery effectively classifies urban LCZs, achieving a test accuracy of 0.75. Temporal factors significantly impact model performance, with higher accuracies observed for daytime (0.8) and Spring/Summer imagery (0.78), as these conditions provide clearer thermal separability for distinguishing LCZs. The model achieved its highest test accuracy (0.83) when trained and tested on thermal images with the highest LST values. This suggests that focusing on high-value LST images with sufficient variability enhances classification performance compared to a generalized approach using the full dataset.

快速城市化对城市微气候构成挑战，使资源紧张，并影响公共卫生。了解微气候动力学是有效减缓和可持续发展的关键。局部气候带（LCZ）分类支持气候适应性规划，但由于城市景观的多样性和复杂性，以及小区域内不同土地利用和材料的共存，使分类变得复杂。虽然LCZ分类通常使用多光谱图像、激光雷达和土地利用数据，但这些来源通常会错过时间热动态模式。热成像卫星图像通过区分结构相似但热行为不同的区域来改进LCZ分类。本研究提出基于深度学习的多时相语义分割，仅基于ECOSTRESS卫星图像的时间热模式对城市lcz进行分类。该方法在荷兰鹿特丹和海牙附近沿海城市的案例研究中得到了应用，并展示了空间和时间因素（昼夜和季节）如何影响语义分割模型在不同LCZ类别上的性能。研究表明，将U-Net架构应用于时空热像图，可以有效地对城市lcz进行分类，测试精度达到0.75。时间因素显著影响模型性能，白天（0.8）和春夏影像（0.78）的观测精度更高，因为这些条件为区分lccs提供了更清晰的热可分性。在LST值最高的热图像上进行训练和测试时，该模型达到了最高的测试精度（0.83）。这表明，与使用完整数据集的广义方法相比，专注于具有足够可变性的高值LST图像可以提高分类性能。

{"title":"Urban local climate zone classification through deep learning using spatio-temporal thermal imagery","authors":"Michaja van Capel , Azarakhsh Rafiee , Roderik Lindenbergh","doi":"10.1016/j.rsase.2026.101889","DOIUrl":"10.1016/j.rsase.2026.101889","url":null,"abstract":"<div><div>Rapid urbanization challenges urban micro-climates, strains resources and affects public health. Understanding micro-climate dynamics is key to effective mitigation and sustainable development. Local Climate Zone (LCZ) classification supports climate-resilient planning but is complicated by the diversity and complexity of diverse urban landscapes and the coexistence of varying land uses and materials within small areas. While LCZ classification typically uses multispectral imagery, LiDAR, and land-use data, these sources often miss temporal thermal dynamic patterns. Thermal satellite imagery improves LCZ classification by distinguishing zones with similar structures but differing thermal behavior. This research proposes using deep learning-based multitemporal semantic segmentation to classify urban LCZs based solely on temporal thermal patterns from ECOSTRESS satellite imagery. The methodology is applied in a in a case study around the near coastal cities of Rotterdam and The Hague in The Netherlands and demonstrates how spatial and temporal factors (both diurnal and seasonal) influence the performance of the semantic segmentation model on different LCZ classes. The study shows that a U-Net architecture applied on spatio-temporal thermal imagery effectively classifies urban LCZs, achieving a test accuracy of 0.75. Temporal factors significantly impact model performance, with higher accuracies observed for daytime (0.8) and Spring/Summer imagery (0.78), as these conditions provide clearer thermal separability for distinguishing LCZs. The model achieved its highest test accuracy (0.83) when trained and tested on thermal images with the highest LST values. This suggests that focusing on high-value LST images with sufficient variability enhances classification performance compared to a generalized approach using the full dataset.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101889"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146077804","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}

引用次数: 0

Lightweight dual-encoder deep learning integrating Sentinel-1 and Sentinel-2 for paddy field mapping 轻量级双编码器深度学习集成Sentinel-1和Sentinel-2稻田测绘

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-02-05 DOI: 10.1016/j.rsase.2026.101895

Bagus Setyawan Wijaya , Rinaldi Munir , Nugraha Priya Utama

Timely and accurate paddy field mapping remains challenging in tropical regions due to persistent cloud cover and complex cropping patterns. We propose DSSNet, a lightweight dual-encoder semantic segmentation framework that fuses Sentinel-1 SAR and Sentinel-2 optical imagery. DSSNet leverages modality-specific backbones from different architectural paradigms: EfficientNet-B0, a convolutional, and MaxVit-T, a transformer-based encoder. To further enhance multimodal feature discrimination, we introduce two axial attention mechanisms — Axial Spatial Attention (ASA) and Axial Channel Attention (ACA) — to selectively emphasize directional spatial patterns and inter-channel relationships. Evaluated on imagery from Indonesia rice-growing regions during the 2019 season, DSSNet achieves an F1-score of 0.8982, pixel accuracy of 0.8998, and mIoU of 0.8156, outperforming ten benchmark models. These findings underscore the operational feasibility of lightweight dual-paradigm fusion architectures for large-scale, in-season agricultural mapping under complex environmental conditions. Our code and model will be publicly available at https://github.com/project4earth/DSSNet.

在热带地区，由于持续的云层覆盖和复杂的种植模式，及时和准确的水田测绘仍然具有挑战性。我们提出了DSSNet，一个轻量级的双编码器语义分割框架，融合了Sentinel-1 SAR和Sentinel-2光学图像。dsnet利用了来自不同架构范例的特定于模式的主干：高效网b0（卷积）和maxvitt（基于转换器的编码器）。为了进一步增强多模态特征识别，我们引入了两种轴向注意机制——轴向空间注意（ASA）和轴向通道注意（ACA），以选择性地强调方向空间模式和通道间关系。通过对2019年印度尼西亚水稻种植区的影像进行评估，DSSNet的f1得分为0.8982，像素精度为0.8998，mIoU为0.8156，优于10个基准模型。这些发现强调了轻量级双范式融合架构在复杂环境条件下用于大规模季节性农业制图的操作可行性。我们的代码和模型将在https://github.com/project4earth/DSSNet上公开提供。

{"title":"Lightweight dual-encoder deep learning integrating Sentinel-1 and Sentinel-2 for paddy field mapping","authors":"Bagus Setyawan Wijaya , Rinaldi Munir , Nugraha Priya Utama","doi":"10.1016/j.rsase.2026.101895","DOIUrl":"10.1016/j.rsase.2026.101895","url":null,"abstract":"<div><div>Timely and accurate paddy field mapping remains challenging in tropical regions due to persistent cloud cover and complex cropping patterns. We propose <strong>DSSNet</strong>, a lightweight dual-encoder semantic segmentation framework that fuses Sentinel-1 SAR and Sentinel-2 optical imagery. DSSNet leverages modality-specific backbones from different architectural paradigms: <em>EfficientNet-B0</em>, a convolutional, and <em>MaxVit-T</em>, a transformer-based encoder. To further enhance multimodal feature discrimination, we introduce two axial attention mechanisms — Axial Spatial Attention (ASA) and Axial Channel Attention (ACA) — to selectively emphasize directional spatial patterns and inter-channel relationships. Evaluated on imagery from Indonesia rice-growing regions during the 2019 season, DSSNet achieves an F1-score of 0.8982, pixel accuracy of 0.8998, and mIoU of 0.8156, outperforming ten benchmark models. These findings underscore the operational feasibility of lightweight dual-paradigm fusion architectures for large-scale, in-season agricultural mapping under complex environmental conditions. Our code and model will be publicly available at <span><span>https://github.com/project4earth/DSSNet</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101895"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147395885","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}

引用次数: 0

Estimation of ICESat-Equivalent Arctic winter sea ice thickness From AMSR-E brightness temperature data based on machine learning approach 基于机器学习方法的AMSR-E亮温数据估算icesat等效北极冬季海冰厚度

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-02-07 DOI: 10.1016/j.rsase.2026.101911

Lin Liu, Lian He, Fengming Hui, Zhuoqi Chen, Xiao Cheng

The Ice, Cloud, and land Elevation Satellite (ICESat) was the first laser altimetry mission to provide estimates of sea ice thickness (SIT) in polar regions. However, ICESat only operated for 2 or 3 months per year due to instrument constraints and provided very sparse observations of SIT in the Arctic Ocean. This study aims to estimate daily pan-Arctic ICESat-equivalent SIT during the wintertime from the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) observed brightness temperatures (TBs) based on an elastic net model (ENM) which was trained using ICESat SIT estimates as the reference. Three types of features were extracted from AMSR-E data, including TBs, spectral gradient ratios (GRs), and polarization ratios (PRs), and their relationships with ICESat SIT estimates were comprehensively investigated. The ENM with the capability of regularization and variable selection was employed for modeling relationship between SIT and microwave features. The SIT estimates were then validated against independent SIT data obtained from moored upward looking sonars (ULS) and Operation IceBridge (OIB) airborne SIT measurements as well as the ENVISAT Radar Altimeter 2 (RA-2) altimetric SIT product. Results suggest that the proposed algorithm could achieve reliable accuracies with root mean square error (RMSE) being about 0.76 m, 0.65 m, and 0.68 m when validating using OIB, ULS and ENVISAT RA-2 data, respectively. More importantly, it successfully captures the seasonal variation of SIT, which allows for the study of spatiotemporal change of Arctic sea ice on daily basis.

冰、云和陆地高程卫星（ICESat）是第一个提供极地海冰厚度（SIT）估计的激光测高任务。然而，由于仪器的限制，ICESat每年只运行2到3个月，并且提供了非常稀疏的北冰洋SIT观测。基于弹性网模型（ENM），以ICESat SIT估算值为参考，利用地球观测系统高级微波扫描辐射计（AMSR-E）观测到的亮度温度（TBs）估算冬季泛北极地区每日ICESat等效SIT。从AMSR-E数据中提取了TBs、光谱梯度比（GRs）和极化比（PRs） 3种特征，并对其与ICESat SIT估算值的关系进行了全面研究。利用具有正则化和变量选择能力的ENM对SIT与微波特征之间的关系进行建模。然后根据独立的SIT数据对SIT估计进行验证，这些数据来自停泊向上看声纳（ULS）和冰桥操作（OIB）机载SIT测量以及ENVISAT雷达高度计2 （RA-2）测高SIT产品。结果表明，该算法在OIB、ULS和ENVISAT RA-2数据验证时，均方误差（RMSE）分别约为0.76 m、0.65 m和0.68 m，精度可靠。更重要的是，它成功地捕获了SIT的季节变化，为研究北极海冰的逐日时空变化提供了基础。

{"title":"Estimation of ICESat-Equivalent Arctic winter sea ice thickness From AMSR-E brightness temperature data based on machine learning approach","authors":"Lin Liu, Lian He, Fengming Hui, Zhuoqi Chen, Xiao Cheng","doi":"10.1016/j.rsase.2026.101911","DOIUrl":"10.1016/j.rsase.2026.101911","url":null,"abstract":"<div><div>The Ice, Cloud, and land Elevation Satellite (ICESat) was the first laser altimetry mission to provide estimates of sea ice thickness (SIT) in polar regions. However, ICESat only operated for 2 or 3 months per year due to instrument constraints and provided very sparse observations of SIT in the Arctic Ocean. This study aims to estimate daily pan-Arctic ICESat-equivalent SIT during the wintertime from the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) observed brightness temperatures (TBs) based on an elastic net model (ENM) which was trained using ICESat SIT estimates as the reference. Three types of features were extracted from AMSR-E data, including TBs, spectral gradient ratios (GRs), and polarization ratios (PRs), and their relationships with ICESat SIT estimates were comprehensively investigated. The ENM with the capability of regularization and variable selection was employed for modeling relationship between SIT and microwave features. The SIT estimates were then validated against independent SIT data obtained from moored upward looking sonars (ULS) and Operation IceBridge (OIB) airborne SIT measurements as well as the ENVISAT Radar Altimeter 2 (RA-2) altimetric SIT product. Results suggest that the proposed algorithm could achieve reliable accuracies with root mean square error (RMSE) being about 0.76 m, 0.65 m, and 0.68 m when validating using OIB, ULS and ENVISAT RA-2 data, respectively. More importantly, it successfully captures the seasonal variation of SIT, which allows for the study of spatiotemporal change of Arctic sea ice on daily basis.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101911"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147395888","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}

引用次数: 0

NDVI-UNet: A novel approach for improved vegetation segmentation using Sentinel-2 images NDVI-UNet：一种利用Sentinel-2图像改进植被分割的新方法

IF 4.5 Q2 ENVIRONMENTAL SCIENCES

Remote Sensing Applications-Society and Environment

Pub Date : 2026-01-01 Epub Date: 2026-01-30 DOI: 10.1016/j.rsase.2026.101905

Fatima Ezahrae Ezzaher , Nizar Ben Achhab , Hafssa Naciri , Naoufal Raissouni

In regions with diverse climates like the Mediterranean, vegetation segmentation is vital for insightful environmental analysis and efficient resource management. Thus, efforts in this field often focus on improving detection accuracy through methods like Vegetation Indices (VIs) from satellite imagery and advanced segmentation techniques. However, both methods face limitations. This study presents a novel approach to mitigate two major drawbacks of these methods: misclassifications of VIs, particularly the blue roof issue presented in the Normalized Difference Vegetation Index (NDVI), and the laborious manual annotation needed to train segmentation models, by merging the two methods and leveraging the strengths of each while mitigating the problems of the other. We analyzed sixteen Sentinel-2 images across four Mediterranean climates and seasons. Vegetation masks were generated using NDVI to train three deep learning models (i.e., UNet, LinkNet, and FPN) with two backbones (i.e., ResNet34 and ResNet50) and three input configurations (i.e., RGB, RGB-NIR, and RGB-NDVI), yielding 18 model combinations. The best-performing model was UNet with ResNet50 and RGB-NDVI, achieving an IoU of 94.82 %, F1-score of 97.28 %, and Accuracy of 98.21 %. We also compared our method with two other automatic labeling techniques: Maximum Entropy and ESA WorldCover map. While both baselines performed well, our method outperformed them with an IoU of 85.87 %, F1-score of 93.05 %, and Accuracy of 91.51 %. Additionally, our approach effectively mitigates the blue roof issue. This study highlights the effectiveness of combining deep learning models with VIs for vegetation segmentation, delivering improved accuracy while significantly lowering the dependence on extensive manual annotation.

在地中海等气候多样的地区，植被分割对于深刻的环境分析和有效的资源管理至关重要。因此，该领域的工作通常集中在通过卫星图像中的植被指数（VIs）和先进的分割技术等方法来提高检测精度。然而，这两种方法都面临局限性。本研究提出了一种新的方法，通过合并两种方法，利用各自的优势，同时减轻另一种方法的问题，来减轻这些方法的两个主要缺点：VIs的错误分类，特别是标准化植被指数（NDVI）中出现的蓝色屋顶问题，以及训练分割模型所需的费力的手工注释。我们分析了横跨四个地中海气候和季节的16张Sentinel-2图像。使用NDVI生成植被掩模，训练三个深度学习模型（即UNet， LinkNet和FPN），具有两个主干（即ResNet34和ResNet50）和三个输入配置（即RGB， RGB- nir和RGB-NDVI），产生18个模型组合。表现最好的UNet模型为ResNet50和RGB-NDVI， IoU为94.82%，f1评分为97.28%，准确率为98.21%。我们还将我们的方法与另外两种自动标记技术：Maximum Entropy和ESA WorldCover map进行了比较。虽然两个基线都表现良好，但我们的方法优于它们，IoU为85.87%，f1评分为93.05%，准确率为91.51%。此外，我们的方法有效地缓解了蓝色屋顶的问题。该研究强调了将深度学习模型与VIs相结合用于植被分割的有效性，提高了准确性，同时显著降低了对大量人工注释的依赖。

{"title":"NDVI-UNet: A novel approach for improved vegetation segmentation using Sentinel-2 images","authors":"Fatima Ezahrae Ezzaher , Nizar Ben Achhab , Hafssa Naciri , Naoufal Raissouni","doi":"10.1016/j.rsase.2026.101905","DOIUrl":"10.1016/j.rsase.2026.101905","url":null,"abstract":"<div><div>In regions with diverse climates like the Mediterranean, vegetation segmentation is vital for insightful environmental analysis and efficient resource management. Thus, efforts in this field often focus on improving detection accuracy through methods like Vegetation Indices (VIs) from satellite imagery and advanced segmentation techniques. However, both methods face limitations. This study presents a novel approach to mitigate two major drawbacks of these methods: misclassifications of VIs, particularly the blue roof issue presented in the Normalized Difference Vegetation Index (NDVI), and the laborious manual annotation needed to train segmentation models, by merging the two methods and leveraging the strengths of each while mitigating the problems of the other. We analyzed sixteen Sentinel-2 images across four Mediterranean climates and seasons. Vegetation masks were generated using NDVI to train three deep learning models (i.e., UNet, LinkNet, and FPN) with two backbones (i.e., ResNet34 and ResNet50) and three input configurations (i.e., RGB, RGB-NIR, and RGB-NDVI), yielding 18 model combinations. The best-performing model was UNet with ResNet50 and RGB-NDVI, achieving an IoU of 94.82 %, F1-score of 97.28 %, and Accuracy of 98.21 %. We also compared our method with two other automatic labeling techniques: Maximum Entropy and ESA WorldCover map. While both baselines performed well, our method outperformed them with an IoU of 85.87 %, F1-score of 93.05 %, and Accuracy of 91.51 %. Additionally, our approach effectively mitigates the blue roof issue. This study highlights the effectiveness of combining deep learning models with VIs for vegetation segmentation, delivering improved accuracy while significantly lowering the dependence on extensive manual annotation.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"41 ","pages":"Article 101905"},"PeriodicalIF":4.5,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147395893","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}

引用次数: 0