[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-5175":3,"related-5175":91},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":18,"tags":20,"primary_topic":9,"secondary_topics":25,"classification":26,"search_phrases":64,"slug":67,"view_count":15,"doi":68,"paper":69,"created_at":90},5175,"Driving forces of contemporary wildfire severity in the American Southwest: a remote sensing and machine learning case study in Arizona, USA","https:\u002F\u002Fdoi.org\u002F10.1080\u002F20964471.2026.2735134","As climate change continues to intensify atmospheric moisture demand across the American Southwest, predicting where wildfires cause severe vegetation mortality remains a major challenge. New thermal and optical harmonized satellite products provide emerging opportunities for evaluating how pre-fire vegetation moisture and water stress influence fire severity compared to other variables. We modeled continuous Relativized differenced Normalized Burn Ratio (RdNBR) for 45 wildfires that burned in the 2019 and 2020 fire seasons using random forest to compare Harmonized Landsat–Sentinel-2 vegetation indices, ECOSTRESS (ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station) evapotranspiration and evaporative-stress metrics, Daymet pre-fire weather variables, LANDFIRE fuels attributes, and topographic predictors across five cumulative pre-ignition windows from 1 week to 3 months before ignition. Despite exceptional drought and anomalous fire-weather conditions, low-severity fire was common over the study period. Integrated models outperformed individual product-family models, indicating that fire severity was best explained by combined antecedent weather, terrain, fuels, vegetation type, and vegetation-moisture predictors. Temperature, elevation, solar radiation, vapor pressure, and HLS Normalized Difference Moisture Index (NDMI) were among the most consistent predictors. NDMI was the most consistent HLS plant-moisture metric, but vegetation moisture alone did not explain fire severity. Canopy cover was also important in woodlands and forests. Models using data from shorter pre-ignition windows outperformed longer pre-fire window models. They varied in performance among vegetation types, with the strongest vegetation-specific performance in forests and woodlands. These findings show that emerging high-resolution satellite products can help characterize pre-fire vegetation conditions and improve fire-risk assessments, especially when integrated with short-term antecedent weather, landscape moisture, and terrain data.","随着气候变化持续加剧美国西南部的大气水分需求，预测野火在何处导致严重植被死亡仍是一项重大挑战。新的热红外与光学协调卫星产品为评估火灾前植被水分和水分胁迫相较于其他变量如何影响火烈度提供了新的机遇。我们利用随机森林对2019年和2020年火季燃烧的45场野火构建了连续相对差分归一化燃烧比（RdNBR）模型，比较了协调陆地卫星—哨兵2号（Harmonized Landsat–Sentinel-2）植被指数、ECOSTRESS（国际空间站生态系统空间热辐射计实验）蒸散发与蒸发胁迫指标、Daymet火前气象变量、LANDFIRE燃料属性以及地形预测因子，并考察了从点火前1周至3个月的五个累积点火前时间窗口。尽管期间出现了极端干旱和异常火险天气条件，研究期内低烈度火仍较为普遍。集成模型的表現优于单一产品类别模型，表明火烈度最好由前期气象、地形、燃料、植被类型和植被水分预测因子的组合来解释。温度、海拔、太阳辐射、水汽压和HLS归一化差异水分指数（NDMI）是最为一致的预测因子。NDMI是最一致的HLS植物水分指标，但仅凭植被水分并不能解释火烈度。冠层覆盖度在疏林和林地中也很重要。使用较短点火前时间窗口数据的模型优于较长火前窗口模型。不同植被类型的模型表现存在差异，其中森林和林地的植被特异性表现最强。这些发现表明，新兴高分辨率卫星产品有助于刻画火前植被状况并改进火险评估，尤其是在与短期前期气象、景观水分和地形数据集成时。",null,"Big Earth Data","2026-10-15T00:00:00Z","论文",10,false,0,{"impact":15,"substance":15,"depth":15,"authority":15,"freshness":15,"relevant":15,"comment":17},"该研究聚焦美国西南部野火严重程度的遥感与机器学习建模，属于生态与灾害遥感领域，与三农、农业信息化、智慧农业等主题无直接关联，不建议进入每日精选。",[19],{"name":10,"url":6},[21,22,23,24],"机器学习","遥感监测","野火风险","植被水分",[],{"status":27,"locked":14,"version":28,"method":29,"evidence":30},"needs_review","2026-10-22.2","rules",{"candidates":31,"agricultural_evidence":14,"reason":63},[32,42,49,55],{"code":33,"score":13,"supported":34,"title":35,"abstract":36,"tags":41},"H7",true,[],[37,38,39,40],"气候变化","干旱","climate change","drought",[23],{"code":43,"score":44,"supported":34,"title":45,"abstract":47,"tags":48},"I2",9,[46],"machine learning",[],[21],{"code":50,"score":44,"supported":34,"title":51,"abstract":53,"tags":54},"I3",[52],"remote sensing",[],[22],{"code":56,"score":57,"supported":34,"title":58,"abstract":59,"tags":62},"H3",6,[],[60,61],"蒸散发","evapotranspiration",[],"农业相关性需要复核",[65,66],"美国西南部 野火 遥感","ECOSTRESS 植被水分 火险","美国西南部野火遥感-5175","10.1080\u002F20964471.2026.2735134",{"doi":68,"openalex_id":70,"authors":71,"venue":10,"cited_by_count":15,"oa_url":82,"card":83,"direction":87,"ingested_from":89},"W7222006390",[72,75,77,79],{"name":73,"orcid":74},"Helen Mills Poulos","https:\u002F\u002Forcid.org\u002F0000-0003-0686-9770",{"name":76,"orcid":9},"Sarah Hardy",{"name":78,"orcid":9},"Jenna Lynne Otaola",{"name":80,"orcid":81},"Andrew  Marder Barton","https:\u002F\u002Forcid.org\u002F0000-0001-5978-0337","https:\u002F\u002Fwww.tandfonline.com\u002Fdoi\u002Fpdf\u002F10.1080\u002F20964471.2026.2735134?needAccess=true",{"tldr":84,"method":85,"finding":86,"direction":87,"opportunity":88},"用遥感与随机森林分析亚利桑那45场野火，识别火前植被水分等驱动因子。","HLS植被指数、ECOSTRESS蒸散、Daymet气象与地形数据，随机森林建模","火险严重度需综合火前短期天气、地形、燃料与植被水分，单一水分指标不足。","农业遥感与作物表型","可借鉴其多源遥感融合与短窗口建模思路，研究作物干旱胁迫与灾损风险预警。","openalex","2026-10-17T23:30:30.060308Z",{"total":57,"page":92,"page_size":57,"items":93},1,[94,195,264,332,430,504],{"id":95,"title":96,"url":97,"summary":98,"summary_zh":99,"content":9,"source_name":100,"source_url":97,"published_at":101,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":102,"score_detail":103,"sources":109,"tags":111,"primary_topic":115,"secondary_topics":121,"classification":133,"search_phrases":171,"slug":174,"view_count":15,"doi":175,"paper":176,"created_at":194},5558,"Estimation of Soil Water Content in Winter Wheat Field at Jointing Stage Using Multi-Source Features and Machine Learning","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fagronomy16202071","Accurate and timely monitoring of soil water content (SWC) plays an essential supporting role in crop irrigation management and smart agricultural practice. Remote sensing technology provides rapid and convenient approaches to large-scale farmland SWC estimation or inversion. However, existing mechanistic modeling methods are limited by parameter localization and land surface heterogeneity, while machine learning data-driven methods lack quite good physical interpretability. To address deficiencies in single data-driven methods and single vegetation microwave scattering models, and explore the estimation effect of different feature variables in machine learning. This study proposes an innovative estimation model that integrates machine learning strategies and multi-source features, including spectral indices and features derived from the classic water cloud model (WCM). The novelty of the proposed model lies in balancing physical interpretability and data-driven accuracy, namely extracting relevant soil backscattering parameters through WCM, which are then used as input features in the machine learning. A combination of field-measured data and unmanned aerial vehicle (UAV) multi-source remote sensing data, including multispectral and synthetic aperture radar (SAR) data, was acquired synchronously at the jointing stage of winter wheat. Through extracting variables such as the SAR microwave backscattering coefficient and the optical vegetation index, and conducting feature screening and parameter sensitivity analysis, we optimized the input variable sets of models based on back-propagation (BP) neural network, support vector regression (SVR) and random forest (RF) algorithms. The study results showed that the established estimation model had excellent estimation performance, especially for the RF algorithm, with the highest R2 of 0.855, lowest MAE of 0.0212 cm3\u002Fcm3, and lowest RMSE of 0.0253 cm3\u002Fcm3. It was also shown that different feature input schemes have obvious effects on SWC estimation performance. On the whole, RF had the best estimation accuracy, followed by SVR and then BPNN. This study demonstrated that models integrating multi-source remote sensing data and machine learning could enhance SWC estimation ability for winter wheat fields, and could also serve as a promising approach for SWC monitoring in winter wheat fields in the North China Plain.","准确、及时地监测土壤含水量（SWC）对作物灌溉管理和智慧农业实践具有重要的支撑作用。遥感技术为大规模农田土壤含水量估算或反演提供了快速便捷的手段。然而，现有机理建模方法受参数本地化和地表异质性限制，机器学习数据驱动方法则缺乏良好的物理解释性。针对单一数据驱动方法和单一植被微波散射模型的不足，并探究不同特征变量在机器学习中的估算效果，本研究提出了一种融合机器学习策略与多源特征的创新估算模型，所采用特征包括光谱指数和经典水云模型（WCM）衍生的特征。该模型的新颖性在于平衡物理解释性与数据驱动精度，即通过WCM提取相关土壤后向散射参数，再将其作为机器学习的输入特征。研究在冬小麦拔节期同步获取了田间实测数据与无人机（UAV）多源遥感数据，包括多光谱和合成孔径雷达（SAR）数据。通过提取SAR微波后向散射系数和光学植被指数等变量，并进行特征筛选和参数敏感性分析，优化了基于反向传播（BP）神经网络、支持向量回归（SVR）和随机森林（RF）算法的模型输入变量集。研究结果表明，所建立的估算模型具有优良的估算性能，尤其是RF算法，最高R²为0.855，最低MAE为0.0212 cm³\u002Fcm³，最低RMSE为0.0253 cm³\u002Fcm³。研究还表明，不同特征输入方案对土壤含水量估算性能有明显影响。总体而言，RF估算精度最佳，SVR次之，BPNN再次之。本研究证明，融合多源遥感数据与机器学习的模型能够提升冬小麦田土壤含水量估算能力，也可为华北平原冬小麦田土壤含水量监测提供一种有前景的方法。","Agronomy","2026-10-21T00:00:00Z",77,{"impact":104,"substance":105,"depth":106,"authority":107,"freshness":13,"relevant":92,"comment":108},16,21,17,13,"该研究融合多源遥感与机器学习估算冬小麦拔节期土壤含水量，方法有创新、结论可靠，对智慧灌溉有参考价值，但属细分领域论文，影响范围有限。",[110],{"name":100,"url":97},[112,21,22,113,114],"智慧农业","土壤墒情","冬小麦",{"code":116,"slug":117,"name":118,"path":119,"parent_code":120},"A2","wheat","小麦","\u002Ftopic\u002Ffield-crops\u002Fwheat","A",[122,128],{"code":123,"slug":124,"name":125,"path":126,"parent_code":127},"H2","soil","土壤健康与改良","\u002Ftopic\u002Fenvironment\u002Fsoil","H",{"code":43,"slug":129,"name":130,"path":131,"parent_code":132},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",{"status":134,"locked":14,"version":28,"method":29,"evidence":135},"classified",{"candidates":136,"agricultural_evidence":34},[137,143,149,153,159,165],{"code":116,"score":138,"supported":34,"title":139,"abstract":141,"tags":142},26,[140,117],"winter wheat",[118,140,117],[114],{"code":123,"score":144,"supported":34,"title":145,"abstract":146,"tags":148},15,[124],[147,124],"土壤",[113],{"code":43,"score":144,"supported":34,"title":150,"abstract":151,"tags":152},[46],[21,46],[21],{"code":50,"score":13,"supported":34,"title":154,"abstract":155,"tags":158},[],[156,52,157],"遥感","multispectral",[22],{"code":56,"score":57,"supported":34,"title":160,"abstract":161,"tags":164},[],[162,163],"灌溉","irrigation",[],{"code":166,"score":167,"supported":34,"title":168,"abstract":169,"tags":170},"I1",4,[],[112],[112],[172,173],"冬小麦 土壤含水量 遥感","多源遥感 机器学习 墒情","冬小麦土壤含水量遥感-5558","10.3390\u002Fagronomy16202071",{"doi":175,"openalex_id":177,"authors":178,"venue":100,"cited_by_count":15,"oa_url":97,"card":188,"direction":193,"ingested_from":89},"W7223156204",[179,182,184,186],{"name":180,"orcid":181},"Mingliang Jiang","https:\u002F\u002Forcid.org\u002F0000-0002-6486-6087",{"name":183,"orcid":9},"Yi Zhang",{"name":185,"orcid":9},"WEI Lei",{"name":187,"orcid":9},"Chengcai Zhang",{"tldr":189,"method":190,"finding":191,"direction":87,"opportunity":192},"融合WCM物理特征与多源遥感，用机器学习估算冬小麦拔节期土壤含水量。","无人机多光谱与SAR数据，WCM参数+光谱指数，BP\u002FSVR\u002FRF建模。","RF最优，R²达0.855，RMSE为0.0253 cm³\u002Fcm³，特征方案影响显著。","可探索物理约束与深度学习耦合，并迁移至多生育期、多作物及不同土壤质地验证。","智慧农业 \u002F 农业物联网","2026-10-22T11:20:18.887301Z",{"id":196,"title":197,"url":198,"summary":199,"summary_zh":200,"content":9,"source_name":201,"source_url":198,"published_at":202,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":203,"sources":205,"tags":207,"primary_topic":210,"secondary_topics":211,"classification":217,"search_phrases":241,"slug":244,"view_count":15,"doi":245,"paper":246,"created_at":263},5485,"Integrating machine learning and geospatial technologies for spatio-temporal analysis of urban expansion and land use transformation in Gambella Town, Ethiopia","https:\u002F\u002Fdoi.org\u002F10.1371\u002Fjournal.pone.0360419","Rapid urbanization and the associated expansion of built-up areas are transforming land resources, ecosystem functions, and the spatial structure of cities. Distinguishing urban growth, urban expansion, and urban sprawl is essential for understanding the processes and patterns of urban transformation and supporting sustainable urban planning. This study integrates machine learning and geospatial technologies to analyze spatio-temporal land use and land cover (LULC) dynamics and urban expansion in Gambella Town, Ethiopia. Multi-temporal Landsat satellite images from 2005, 2015, and 2025 were processed using remote sensing techniques and a supervised machine learning approach based on the Random Forest (RF) classifier. Six major LULC classes were identified: Agricultural Land, Bush with Grass, Built-Up Area, Bare Land, Forest Land, and Water Bodies. The classification achieved overall accuracies of 90%, 90%, and 89% for 2005, 2015, and 2025, respectively. The results revealed substantial land transformation and rapid physical urban expansion, with the built-up area increasing from 93 ha (2.0%) in 2005–1,956 ha (41.3%) in 2025. This expansion was accompanied by major transformations of surrounding land resources, particularly the conversion of Bush with Grass and other open areas to built-up land. Agricultural land increased from 634 ha (13.4%) to 1,750 ha (36.9%), whereas Bush with Grass declined from 3,375 ha (71.1%) to 652 ha (13.7%). Socio-economic information collected from 277 respondents indicated that population growth, rural–urban migration, housing demand, employment opportunities, and infrastructure development were major drivers of urban expansion. The study distinguishes urban expansion, represented by the increase in built-up land, from urban sprawl, which refers to the spatial pattern and form of outward urban development. The integration of Random Forest classification and GIS-based spatial analysis provides valuable evidence of land transformation, vegetation loss, and increasing pressure on urban services, supporting more sustainable urban planning and land management in Gambella Town.","快速城市化及与之相关的建成区扩张正在改变土地资源、生态系统功能和城市空间结构。区分城市增长（urban growth）、城市扩张（urban expansion）和城市蔓延（urban sprawl），对于理解城市转型的过程与格局、支撑可持续城市规划至关重要。本研究融合机器学习与地理空间技术，分析埃塞俄比亚甘贝拉镇（Gambella Town）土地利用与土地覆盖（LULC）的时空动态及城市扩张。采用遥感技术和基于随机森林（RF）分类器的监督机器学习方法，处理了2005年、2015年和2025年的多时相Landsat卫星影像。研究识别出六类主要LULC类型：农业用地、灌草地、建成区、裸地、林地和水体。2005年、2015年和2025年的分类总体精度分别为90%、90%和89%。结果表明，土地发生了显著转变，城市实体空间快速扩张，建成区面积从2005年的93公顷（2.0%）增至2025年的1,956公顷（41.3%）。这一扩张伴随着周边土地资源的重大转变，尤其是灌草地及其他开阔地被转化为建成用地。农业用地从634公顷（13.4%）增至1,750公顷（36.9%），而灌草地则从3,375公顷（71.1%）降至652公顷（13.7%）。从277名受访者收集的社会经济信息表明，人口增长、城乡迁移、住房需求、就业机会和基础设施建设是城市扩张的主要驱动因素。本研究将城市扩张（以建成用地增加为表征）与城市蔓延（指城市向外发展的空间格局与形态）加以区分。随机森林分类与基于GIS的空间分析的融合，为土地转变、植被丧失及城市服务压力加剧提供了有价值的证据，有助于甘贝拉镇实现更可持续的城市规划与土地管理。","PLoS ONE","2026-10-20T00:00:00Z",{"impact":15,"substance":15,"depth":15,"authority":15,"freshness":15,"relevant":15,"comment":204},"研究聚焦埃塞俄比亚城市扩张与土地利用变化，属城市地理与遥感应用，与三农、农业信息化主题无直接关联，不建议进入每日精选。",[206],{"name":201,"url":198},[21,22,208,209],"土地利用","城市扩张",{"code":43,"slug":129,"name":130,"path":131,"parent_code":132},[212],{"code":213,"slug":214,"name":215,"path":216,"parent_code":132},"I4","gis","GIS与空间分析","\u002Ftopic\u002Fdigital-agriculture\u002Fgis",{"status":134,"locked":14,"version":28,"method":29,"evidence":218},{"candidates":219,"agricultural_evidence":34},[220,224,231,236],{"code":43,"score":144,"supported":34,"title":221,"abstract":222,"tags":223},[46],[21,46],[21],{"code":213,"score":44,"supported":34,"title":225,"abstract":226,"tags":230},[],[227,228,229],"空间分析","GIS","spatial analysis",[],{"code":50,"score":232,"supported":34,"title":233,"abstract":234,"tags":235},7,[],[156,52],[22],{"code":237,"score":92,"supported":14,"title":238,"abstract":239,"tags":240},"H6",[],[],[208],[242,243],"Gambella Town 城市扩张","Random Forest 土地利用分类","GambellaTown城市扩张-5485","10.1371\u002Fjournal.pone.0360419",{"doi":245,"openalex_id":247,"authors":248,"venue":201,"cited_by_count":15,"oa_url":198,"card":258,"direction":87,"ingested_from":89},"W7223022155",[249,252,254,256],{"name":250,"orcid":251},"Kanenus Fufa Dararo","https:\u002F\u002Forcid.org\u002F0009-0007-5747-1558",{"name":253,"orcid":9},"Goanar Jal Dup",{"name":255,"orcid":9},"Weyesa Merga Chemeda",{"name":257,"orcid":9},"Shankar Karuppannan",{"tldr":259,"method":260,"finding":261,"direction":87,"opportunity":262},"用机器学习与遥感分析埃塞俄比亚甘贝拉镇2005-2025年城市扩张与土地利用变化。","Landsat多时相影像、随机森林分类与GIS空间分析，结合277份问卷。","建成区从93公顷增至1956公顷，灌草地大幅转为建成区和农地。","可将随机森林时序分类与驱动力调查结合，扩展到更多非洲小城镇并耦合城市增长模拟。","2026-10-21T23:30:23.271293Z",{"id":265,"title":266,"url":267,"summary":268,"summary_zh":269,"content":9,"source_name":270,"source_url":267,"published_at":271,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":272,"score_detail":273,"sources":278,"tags":280,"primary_topic":9,"secondary_topics":284,"classification":285,"search_phrases":297,"slug":300,"view_count":15,"doi":301,"paper":302,"created_at":331},4886,"Mapping Fuel Categories in Mediterranean Regions with Sentinel-2 and PlanetScope Satellite Acquisitions","https:\u002F\u002Fdoi.org\u002F10.5194\u002Fisprs-archives-xlviii-m-13-2026-9-2026","Accurate and spatially explicit mapping of forest fuel types is essential for developing integrated wildfire risk management strategies, estimating fire danger, and modeling fire propagation in fire-prone Mediterranean ecosystems. While time series of Copernicus Sentinel-2 multispectral imagery at 10 m spatial resolution have proven effective for regional fuel mapping, severe spectral and spatial ambiguities persist in distinguishing complex canopy structures, heterogeneous shrublands, and pastures. This study explores the synergistic integration of Sentinel-2 time series and high-resolution PlanetScope satellite acquisitions (3 m spatial resolution) to produce a 10 m fuel type map (9 classes) over a pilot area in Sardinia, Italy. A total of 114 features were synthesized, combining dynamic Sentinel-2 metrics (kNDVI, FAPAR, NDMI, ReNDVI harmonics), PlanetScope seasonal kNDVI zonal statistics and Gray-Level Co-occurrence Matrix (GLCM) textural features, and static terrain\u002Fcanopy parameters. Using an XGBoost machine learning classifier trained on 245 photo-interpreted Regions of Interest (80\u002F20 spatial split), three sets of features were compared. The combined model achieved a test overall accuracy of 74.46%, compared to 74.05% for Sentinel-2 alone and 74.66% for PlanetScope alone. Notably, incorporating PlanetScope textural features significantly improved classification metrics for dense canopy classes, increasing the F1-score of broadleaf evergreen trees and needleleaf evergreen trees. The PlanetScope GLCM kNDVI autumn entropy emerged as the most important feature overall. Conversely, PlanetScope performed poorly on the grass class. These findings highlight the critical trade-offs and high complementarity between fine spatial texture and broad spectral coverage for operational wildfire management.","在易发生火灾的地中海生态系统中，准确且具有空间明确性的森林可燃物类型制图对于制定综合野火风险管理策略、估算火险以及模拟火蔓延至关重要。尽管10 m空间分辨率的哥白尼Sentinel-2多光谱影像时间序列已被证明对区域可燃物制图有效，但在区分复杂冠层结构、异质性灌丛和草地时，仍存在严重的光谱和空间模糊性。本研究探索了Sentinel-2时间序列与高分辨率PlanetScope卫星影像（3 m空间分辨率）的协同集成，以在意大利撒丁岛的一个试验区生成10 m可燃物类型图（9类）。共合成114个特征，结合了动态Sentinel-2指标（kNDVI、FAPAR、NDMI、ReNDVI谐波）、PlanetScope季节性kNDVI分区统计和灰度共生矩阵（GLCM）纹理特征，以及静态地形\u002F冠层参数。使用在245个照片判读感兴趣区（80\u002F20空间划分）上训练的XGBoost机器学习分类器，比较了三组特征。组合模型的测试总体精度为74.46%，而仅使用Sentinel-2为74.05%，仅使用PlanetScope为74.66%。值得注意的是，纳入PlanetScope纹理特征显著改善了密集冠层类的分类指标，提高了阔叶常绿乔木和针叶常绿乔木的F1分数。PlanetScope GLCM kNDVI秋季熵成为总体最重要的特征。相反，PlanetScope在草地类上表现较差。这些发现凸显了精细空间纹理与广域光谱覆盖在业务化野火管理中的关键权衡和高互补性。","˜The œinternational archives of the photogrammetry, remote sensing and spatial information sciences\u002FInternational archives of the photogrammetry, remote sensing and spatial information sciences","2026-10-12T00:00:00Z",72,{"impact":274,"substance":275,"depth":106,"authority":107,"freshness":276,"relevant":92,"comment":277},12,22,8,"方法新颖、数据扎实的遥感制图论文，对林火风险管理有参考价值，但属细分领域研究，公共影响有限。",[279],{"name":270,"url":267},[281,21,22,282,283],"Sentinel-2","地中海农业","林火风险",[],{"status":27,"locked":14,"version":28,"method":29,"evidence":286},{"candidates":287,"agricultural_evidence":14,"reason":63},[288,292],{"code":43,"score":232,"supported":34,"title":289,"abstract":290,"tags":291},[],[21,46],[21],{"code":50,"score":293,"supported":34,"title":294,"abstract":295,"tags":296},5,[],[157],[281,22],[298,299],"Sentinel-2 PlanetScope 燃料类型制图","撒丁岛 林火燃料 遥感","Sentinel-2PlanetScope燃料类型制图-4886","10.5194\u002Fisprs-archives-xlviii-m-13-2026-9-2026",{"doi":301,"openalex_id":303,"authors":304,"venue":270,"cited_by_count":15,"oa_url":325,"card":326,"direction":193,"ingested_from":89},"W7221532579",[305,308,311,313,316,319,322],{"name":306,"orcid":307},"Federico Filipponi","https:\u002F\u002Forcid.org\u002F0000-0002-1419-3734",{"name":309,"orcid":310},"Daniela Stroppiana","https:\u002F\u002Forcid.org\u002F0000-0002-5619-4305",{"name":312,"orcid":9},"Forough Rajabi",{"name":314,"orcid":315},"Debora Voltolina","https:\u002F\u002Forcid.org\u002F0000-0001-9186-0644",{"name":317,"orcid":318},"Bachisio Arca","https:\u002F\u002Forcid.org\u002F0000-0002-2834-2753",{"name":320,"orcid":321},"Michele Salis","https:\u002F\u002Forcid.org\u002F0000-0002-0499-9726",{"name":323,"orcid":324},"Francesco Zucca","https:\u002F\u002Forcid.org\u002F0000-0003-4119-5484","https:\u002F\u002Fisprs-archives.copernicus.org\u002Farticles\u002FXLVIII-M-13-2026\u002F9\u002F2026\u002Fisprs-archives-XLVIII-M-13-2026-9-2026.pdf",{"tldr":327,"method":328,"finding":329,"direction":87,"opportunity":330},"融合Sentinel-2时序与PlanetScope高分影像，用XGBoost绘制地中海9类可燃物类","Sentinel-2时序谐波、PlanetScope kNDVI与GLCM纹理及","联合模型精度74.46%，PlanetScope纹理显著提升密冠层F1，但草地类表现差。","可探索多源高分纹理与光谱互补机制，优化稀疏植被与草地类别的特征设计。","2026-10-14T23:30:11.734574Z",{"id":333,"title":334,"url":335,"summary":336,"summary_zh":337,"content":9,"source_name":338,"source_url":335,"published_at":339,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":340,"score_detail":341,"sources":346,"tags":348,"primary_topic":351,"secondary_topics":355,"classification":361,"search_phrases":405,"slug":408,"view_count":409,"doi":410,"paper":411,"created_at":429},4575,"A review of remote sensing for agricultural drought: technologies, indices, modelling advances and future prospects","https:\u002F\u002Fdoi.org\u002F10.1080\u002F15481603.2026.2738981","Agricultural drought is a complex hazard that jeopardises food security and progress toward the Sustainable Development Goals. We systematically reviewed peer-reviewed English-language literature published between 2015 and 2025, focusing on Remote Sensing (RS)-driven agricultural drought monitoring and prediction. Eligible studies were identified from the Web of Science and Scopus databases (last searched in August 2025); review articles, duplicates, and irrelevant studies were excluded, resulting in 181 articles. This review critically examined the evolution of sensors, drought indices, modelling approaches, and emerging technologies used to assess agricultural drought across diverse agro-ecological regions. Specifically, we (1) systematically evaluated the use of optical, thermal, and microwave RS platforms, along with satellite-based, model-based, and reanalysis products for agricultural drought monitoring; (2) reviewed widely used RS-based drought indices, including single indices grouped into nine categories, as well as composite indices; (3) reviewed software or programming languages applied; (4) assessed the progression of modelling approaches from traditional statistical and time-series methods to machine learning (ML) and deep learning (DL) frameworks; and (5) examined validation strategies applied to agricultural drought assessment. The analysis reveals a growing research interest in this topic, with a strong geographic concentration in Asia, particularly China. Optical and thermal platforms (e.g. MODIS and Landsat) dominate the literature (~54%) due to their accessibility, followed by satellite-derived precipitation products (17%), while microwave soil-moisture missions remain underutilized (~7%). Regarding RS-based indices, most studies employed more than one index: 44% analysed multiple individual indices, 51% used composite indices (either alone or in combination with individual indices), and only 5% relied on a single index. Vegetation indices, such as NDVI, remain workhorses; however, their sensitivity to soil background effects and canopy saturation has driven increased adoption of thermal composites (e.g. VHI), as well as adaptive weighting schemes and principal component analysis-based approaches. Moreover, computational tools used in drought research can be broadly grouped into geospatial and RS platforms for large-scale data processing and analysis (e.g. QGIS, GEE), and general-purpose programming languages (e.g. R, Python) for model development and advanced analytics. Regarding the methodology, statistical and time-series approaches account for 64% of the reviewed studies, followed by ML models (20%) and DL methods (15%), while explainable artificial intelligence (XAI) remains nascent. Techniques, such as random forests, support vector machines, and CNN and LSTM architectures, generally outperform traditional models but require large training datasets and raise concerns related to interpretability and ethics. In terms of validation, hydro-climatic indices such as SPI and SPEI are most commonly used (32%), followed by crop yield and yield anomaly data (19%). This review further identifies key research gaps across sensors, indices, and modelling approaches and proposes a roadmap for developing equitable, interpretable, and resilient agricultural drought monitoring systems. Emerging opportunities include context-aware and interpretable AI, multi-sensor and multi-scale data fusion across optical, thermal, and microwave domains, integration of physiologically meaningful indicators such as sun-induced chlorophyll fluorescence (SIF), robust uncertainty quantification, and the development of hybrid and physics-informed modelling frameworks.","农业干旱是一种复杂的灾害，威胁粮食安全和可持续发展目标的实现。我们系统综述了2015年至2025年间发表的同行评议英文文献，聚焦于遥感（RS）驱动的农业干旱监测与预测。符合条件的研究通过Web of Science和Scopus数据库检索确定（最后检索时间为2025年8月）；综述文章、重复文献及不相关研究被排除，最终纳入181篇文献。本综述批判性地审视了用于评估不同农业生态区农业干旱的传感器、干旱指数、建模方法及新兴技术的演变。具体而言，我们（1）系统评估了光学、热红外和微波遥感平台，以及基于卫星、基于模型和再分析产品在农业干旱监测中的应用；（2）综述了广泛使用的基于遥感的干旱指数，包括归为九类的单一指数以及复合指数；（3）综述了所应用的软件或编程语言；（4）评估了建模方法从传统统计和时间序列方法向机器学习（ML）和深度学习（DL）框架的演进；（5）考察了应用于农业干旱评估的验证策略。分析显示，该主题的研究兴趣日益增长，地理上高度集中于亚洲，尤其是中国。光学和热红外平台（如MODIS和Landsat）因其可获取性在文献中占主导地位（约54%），其次是卫星降水产品（17%），而微波土壤水分任务仍未得到充分利用（约7%）。在基于遥感的指数方面，大多数研究采用了不止一种指数：44%分析了多个单一指数，51%使用了复合指数（单独使用或与单一指数结合使用），仅5%依赖单一指数。植被指数（如NDVI）仍是主力工具；然而，其对土壤背景效应和冠层饱和的敏感性推动了热红外复合指数（如VHI）以及自适应权重方案和基于主成分分析的方法得到更多采用。此外，干旱研究中使用的计算工具可大致分为用于大规模数据处理和分析的地理空间与遥感平台（如QGIS、GEE），以及用于模型开发的通用编程语言（如R、Python）。","GIScience & Remote Sensing","2026-10-08T00:00:00Z",81,{"impact":342,"substance":275,"depth":343,"authority":344,"freshness":276,"relevant":92,"comment":345},18,19,14,"系统综述181篇文献，梳理遥感干旱监测的传感器、指数与建模进展，方法学价值高，但属学术综述，产业落地影响有限。",[347],{"name":338,"url":335},[112,349,21,22,350],"粮食安全","农业干旱",{"code":50,"slug":352,"name":353,"path":354,"parent_code":132},"remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing",[356,360],{"code":33,"slug":357,"name":358,"path":359,"parent_code":127},"climate","气候适应与防灾减灾","\u002Ftopic\u002Fenvironment\u002Fclimate",{"code":43,"slug":129,"name":130,"path":131,"parent_code":132},{"status":134,"locked":14,"version":28,"method":29,"evidence":362},{"candidates":363,"agricultural_evidence":34},[364,368,372,379,385,389,396,401],{"code":50,"score":144,"supported":34,"title":365,"abstract":366,"tags":367},[52],[156,52],[22],{"code":33,"score":344,"supported":34,"title":369,"abstract":370,"tags":371},[40],[38,40],[],{"code":43,"score":13,"supported":34,"title":373,"abstract":374,"tags":378},[],[375,21,46,376,377],"深度学习","deep learning","artificial intelligence",[21],{"code":380,"score":232,"supported":34,"title":381,"abstract":382,"tags":384},"A8",[],[349,383],"food security",[349],{"code":123,"score":57,"supported":34,"title":386,"abstract":387,"tags":388},[],[147,124],[],{"code":390,"score":391,"supported":34,"title":392,"abstract":393,"tags":395},"D",3,[],[394],"水产",[],{"code":237,"score":391,"supported":34,"title":397,"abstract":398,"tags":400},[],[399],"农业生态",[],{"code":166,"score":92,"supported":14,"title":402,"abstract":403,"tags":404},[],[],[112],[406,407],"遥感 农业干旱 监测","GIScience Remote Sensing 干旱指数","遥感农业干旱监测-4575",2,"10.1080\u002F15481603.2026.2738981",{"doi":410,"openalex_id":412,"authors":413,"venue":338,"cited_by_count":15,"oa_url":423,"card":424,"direction":193,"ingested_from":89},"W7221081979",[414,417,420],{"name":415,"orcid":416},"Fatemeh Saba","https:\u002F\u002Forcid.org\u002F0000-0001-6171-155X",{"name":418,"orcid":419},"Nima Ahmadian","https:\u002F\u002Forcid.org\u002F0000-0003-3676-1840",{"name":421,"orcid":422},"Olena Dubovyk","https:\u002F\u002Forcid.org\u002F0000-0002-7338-3167","https:\u002F\u002Fwww.tandfonline.com\u002Fdoi\u002Fpdf\u002F10.1080\u002F15481603.2026.2738981?needAccess=true",{"tldr":425,"method":426,"finding":427,"direction":87,"opportunity":428},"系统综述2015-2025年遥感农业干旱监测与预测的传感器、指数、建模及验证进展。","基于Web of Science与Scopus的181篇文献系统综述，涵盖光学\u002F","光学热红外平台占主导，微波土壤水分利用不足；复合指数与机器学习建模成为趋势。","微波土壤水分数据融合与深度学习干旱预测模型在数据稀缺区验证是潜在研究空白。","2026-10-10T23:30:14.933078Z",{"id":431,"title":432,"url":433,"summary":434,"summary_zh":435,"content":9,"source_name":436,"source_url":433,"published_at":437,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":438,"score_detail":439,"sources":441,"tags":443,"primary_topic":447,"secondary_topics":448,"classification":452,"search_phrases":479,"slug":482,"view_count":92,"doi":483,"paper":484,"created_at":503},4527,"Socio‐Economic Impact of Sustainable Development of Water Resources in the Semi‐Arid Regions of Village Raleras, Maharashtra: A Case Study Using Machine Learning Techniques","https:\u002F\u002Fdoi.org\u002F10.1111\u002F1477-8947.70100","ABSTRACT From a socio‐economic perspective, it is of interest to find out whether sustainable water resources can truly sustain rural livelihoods. Based on historical research, it is known that an improvement on resilient economic indicators such as ground water is often correlated with socio‐economic improvements and other aspects of rural livelihoods. The extent of factors affecting livelihood sustenance and improvement in the village of Raleras, Maharashtra, aids in creating a definite understanding of the facets correlating with groundwater restoration. Subsistence channels, arising as an outcome of the created impoundment structures, are tested for their impact on income gains and livelihood improvement. Examination on collected opinion was carried out to provide insights into the ‘before’ and ‘after’ (2015–2023) conditions. A five‐capitals framework has been employed for the interpretation of these results, along with a multivariate analysis to predict a causal framework for the improvement of livelihoods in agrarian economies through creation of check dams and watershed management. Time series study shows that there is greater variance in the NDVI progressively from 2021 to 2023 with values ranging from 0.25 to 0.50 in 2015, to 0.50–0.75 by 2021. Further increase in green patches in 2023 indicates enhanced vegetation health especially in the northern regions. Alternately, NDWI values elaborate initial severe water stress (−0.25 to 0.00) with progressive improvements in localized areas. The spatial variation map shows that eventually there is a notable improvement with more land supporting shrubbery along with more water bodies. The prominent increase in vegetation particularly in the later years can be attributed to water management initiatives undertaken in Raleras. The distributions of the annual average revenue from agricultural operations in 2015 (M = 146,907, SD = 95445.8) and 2023 (M = 706860.5, SD = 655748.3) showed a statistically significant improvement [ t (214) = 13.06, p \u003C 0.001]. The low p ‐value of 0.001205 ( p \u003C 0.05) of the educational facilities utilised indicates a statistically significant result, proving the association between educational quality and improved water facilities. All of the variables under evaluation, including the respondents' vehicle type, landholding type, and house type, have large and positive chi‐squared values, suggesting that the ‘before’ and ‘after’ distributions for each of these variables differ. Principal Component Analysis is used to summarise correlated livelihood indicators and reveals that income growth and livelihood improvement is multi‐dimensional. Multiple pathways to income change are identified through Partial Least Squares Regression, including agriculture and market‐linked channels. A Random Forest Regression Model is utilised to assess the relative importance of livelihood channels and baseline characteristics. Results indicate that improvement in irrigation access, crop diversification, housing quality, livestock ownership and female literacy are strongly associated with higher income gains. Hence, it may be said that the groundwater recharge as a result of construction of check dams must have improved agricultural activity or agricultural output in a material manner, and accordingly all capitals for Raleras have improved.","从社会经济视角来看，探究可持续水资源能否真正维持农村生计具有重要意义。基于历史研究可知，地下水等韧性经济指标的改善往往与社会经济进步及农村生计的其他方面相关联。马哈拉施特拉邦拉莱拉斯村影响生计维持与改善的因素程度，有助于明确理解与地下水恢复相关的各个维度。作为所建蓄水设施产物的生计渠道，其对面收入增长和生计改善的影响得到了检验。对收集到的意见进行了考察，以深入了解“之前”和“之后”（2015—2023年）的状况。研究采用了五资本框架来解读这些结果，并结合多变量分析，预测通过修建拦水坝和流域管理来改善农业经济生计的因果框架。时间序列研究表明，从2021年到2023年，NDVI的变异性逐渐增大，2015年数值范围为0.25至0.50，到2021年升至0.50至0.75。2023年绿色斑块进一步增加，表明植被健康状况改善，尤其是在北部地区。另一方面，NDWI值显示初期存在严重的水分胁迫（−0.25至0.00），后在局部地区逐步改善。空间变异图显示，最终出现了显著改善，更多土地能够支持灌木生长，同时水体也有所增加。植被的显著增加，尤其是在后期，可归因于拉莱拉斯村所实施的水资源管理举措。2015年（M = 146,907，SD = 95445.8）和2023年（M = 706860.5，SD = 655748.3）农业经营年均收入的分布显示出统计学上显著的改善[t(214) = 13.06，p \u003C 0.001]。所利用教育设施的p值为0.001205（p \u003C 0.05），表明结果具有统计学显著性，证明了教育质量与改善的供水设施之间的关联。所有受评估变量，包括受访者的车辆类型、土地持有类型和房屋类型，均具有较大且正向的卡方值，表明这些变量各自的“之前”和“之后”分布存在差异。","Natural Resources Forum","2026-10-07T00:00:00Z",73,{"impact":274,"substance":275,"depth":342,"authority":107,"freshness":276,"relevant":92,"comment":440},"以机器学习与遥感数据量化印度半干旱村庄水资源可持续性对生计的影响，方法扎实、结论具体，但属境外案例研究，对国内三农信息化仅有方法借鉴价值。",[442],{"name":436,"url":433},[444,21,445,22,446],"乡村振兴","水资源管理","灌溉农业",{"code":43,"slug":129,"name":130,"path":131,"parent_code":132},[449],{"code":56,"slug":163,"name":450,"path":451,"parent_code":127},"水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation",{"status":134,"locked":14,"version":28,"method":29,"evidence":453},{"candidates":454,"agricultural_evidence":34},[455,459,464,470,475],{"code":43,"score":44,"supported":34,"title":456,"abstract":457,"tags":458},[46],[],[21],{"code":56,"score":276,"supported":34,"title":460,"abstract":461,"tags":463},[],[462,163],"水资源",[445,446],{"code":465,"score":391,"supported":34,"title":466,"abstract":467,"tags":469},"C",[],[468],"livestock",[],{"code":471,"score":92,"supported":14,"title":472,"abstract":473,"tags":474},"G",[],[],[444],{"code":50,"score":92,"supported":14,"title":476,"abstract":477,"tags":478},[],[],[22],[480,481],"Raleras 马哈拉施特拉 水资源","check dam 地下水 生计","Raleras马哈拉施特拉水资源-4527","10.1111\u002F1477-8947.70100",{"doi":483,"openalex_id":485,"authors":486,"venue":436,"cited_by_count":15,"oa_url":9,"card":496,"direction":502,"ingested_from":89},"W7220922038",[487,490,492,494],{"name":488,"orcid":489},"Papiya Deb","https:\u002F\u002Forcid.org\u002F0000-0002-6554-5111",{"name":491,"orcid":9},"Roopsha Deb",{"name":493,"orcid":9},"Udaysinh Bhosale",{"name":495,"orcid":9},"Vemula Sai Vikas",{"tldr":497,"method":498,"finding":499,"direction":500,"opportunity":501},"用机器学习评估印度半干旱村庄水资源可持续开发对农村生计的社会经济影响。","NDVI\u002FNDWI遥感、五资本框架、多元分析与卡方检验，2015–2023数据。","水管理举措显著提升植被覆盖与农业收入，教育设施改善与水资源改善显著相关。","数字乡村与农业信息化","可结合遥感时序与农户调查，量化水利工程对生计的因果效应并推广至其他半干旱区。","农业人工智能与决策模型","2026-10-09T23:30:50.079298Z",{"id":505,"title":506,"url":507,"summary":508,"summary_zh":509,"content":9,"source_name":510,"source_url":507,"published_at":511,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":512,"score_detail":513,"sources":515,"tags":517,"primary_topic":9,"secondary_topics":521,"classification":522,"search_phrases":534,"slug":537,"view_count":15,"doi":538,"paper":539,"created_at":556},4519,"Predicting water quality in mine-affected environments: A hyperspectral remote sensing study in the Iberian pyrite Belt","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.srs.2026.100516","Monitoring water quality in Acid Mine Drainage (AMD) environments requires high-resolution data to capture the complex interplay between geochemical and hydrological dynamics. This study integrates unmanned aerial vehicle (UAV)-based hyperspectral imaging with machine learning to estimate pH, sulfate (SO 4 2- ), and iron (Fe) concentrations across nine water bodies at the Tharsis and Lagunazo abandoned mining sites in the Iberian Pyrite Belt (SW Iberian Peninsula), which span a broad hydrochemical gradient. Using in situ measurements for training, five algorithms: Random Forest (RF), k-nearest neighbors (KNN), Partial Least Squares Regression (PLSR), XGBoost, and Multilayer Perceptron (MLP), against dimensionality reduction techniques (LASSO and PCA) were evaluated. The MLP model using the full spectral set yielded the best performance for pH (R 2 = 0.952), while RF models using LASSO-selected features performed best for Fe (R 2 = 0.861) and SO 4 2- (R 2 = 0.880). SHapley Additive exPlanations (SHAP) confirmed that model decisions align with known spectral absorption features. Building on previous hyperspectral AMD studies, this work advances the state of the art by (i) exploiting UAV-scale hyperspectral data to resolve fine-scale spatial variability within individual water bodies, (ii) explicitly distinguishing between optically active and non-optically active water quality parameters, and (iii) systematically comparing multiple machine learning regression algorithms to retrieve pH, iron, and sulfate across chemically diverse systems. By integrating high-resolution imagery with hydrochemical data, this study provides a robust framework for monitoring AMD-impacted surface waters at spatial scales relevant for environmental management.","在酸性矿山排水（AMD）环境中监测水质需要高分辨率数据，以捕捉地球化学过程与水文动态之间复杂的相互作用。本研究将基于无人机（UAV）的高光谱成像与机器学习相结合，估算了伊比利亚黄铁矿带（伊比利亚半岛西南部）Tharsis和Lagunazo废弃矿区九个水体的pH、硫酸盐（SO₄²⁻）和铁（Fe）浓度，这些水体跨越了广泛的水化学梯度。利用原位测量数据进行训练，评估了五种算法：随机森林（RF）、k近邻（KNN）、偏最小二乘回归（PLSR）、XGBoost和多层感知机（MLP），并结合降维技术（LASSO和PCA）进行比较。使用全光谱集的MLP模型在pH估算中表现最佳（R² = 0.952），而使用LASSO筛选特征的RF模型在Fe（R² = 0.861）和SO₄²⁻（R² = 0.880）估算中表现最优。SHapley加法解释（SHAP）证实模型决策与已知的光谱吸收特征相一致。在以往高光谱AMD研究的基础上，本工作通过以下方面推进了该领域的技术水平：（i）利用无人机尺度的高光谱数据解析单个水体内部的精细空间变异性，（ii）明确区分光学活性与非光学活性水质参数，以及（iii）系统比较多种机器学习回归算法以反演化学组成多样系统中的pH、铁和硫酸盐。通过将高分辨率影像与水化学数据相结合，本研究为在与环境管理相关的空间尺度上监测受AMD影响的地表水提供了一个稳健的框架。","Science of Remote Sensing","2026-10-06T00:00:00Z",78,{"impact":342,"substance":275,"depth":342,"authority":344,"freshness":57,"relevant":92,"comment":514},"该研究将无人机高光谱与机器学习结合用于酸性矿山废水水质反演，方法新颖、数据扎实，对矿区农业水环境监测具有参考价值，但属细分领域国际论文，公共影响有限。",[516],{"name":510,"url":507},[21,518,22,519,520],"无人机遥感","水质监测","矿山生态修复",[],{"status":27,"locked":14,"version":28,"method":29,"evidence":523},{"candidates":524,"agricultural_evidence":14,"reason":63},[525,530],{"code":50,"score":105,"supported":34,"title":526,"abstract":528,"tags":529},[52,527],"hyperspectral",[527],[518,22],{"code":43,"score":232,"supported":34,"title":531,"abstract":532,"tags":533},[],[21,46],[21],[535,536],"伊比利亚黄铁矿带 酸性矿山废水 高光谱","UAV 高光谱 水质反演","伊比利亚黄铁矿带酸性矿山废水高光谱-4519","10.1016\u002Fj.srs.2026.100516",{"doi":538,"openalex_id":540,"authors":541,"venue":510,"cited_by_count":15,"oa_url":550,"card":551,"direction":87,"ingested_from":89},"W7220716726",[542,544,547],{"name":543,"orcid":9},"Melisa Isgró",{"name":545,"orcid":546},"Maria  Dolores Basallote","https:\u002F\u002Forcid.org\u002F0000-0003-2011-3806",{"name":548,"orcid":549},"L. Barbero","https:\u002F\u002Forcid.org\u002F0000-0002-3513-2025","https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fpii\u002FS2666017226001549\u002Fpdf",{"tldr":552,"method":553,"finding":554,"direction":87,"opportunity":555},"用无人机高光谱与机器学习反演矿区酸性废水pH、硫酸盐和铁浓度。","无人机高光谱成像+5种机器学习算法，结合LASSO\u002FPCA降维与SHAP解释。","MLP对pH最佳(R²=0.952)，RF结合LASSO对Fe和SO₄²⁻最佳。","可将该高光谱-机器学习框架迁移到农业面源污染水体监测，并探索跨区域迁移学习。","2026-10-09T23:30:30.251656Z"]