[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2213":3,"related-2213":107},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":8,"source_name":9,"source_url":8,"published_at":10,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":14,"score_detail":15,"sources":23,"tags":25,"primary_topic":32,"secondary_topics":38,"classification":39,"search_phrases":90,"slug":93,"view_count":94,"doi":95,"paper":96,"created_at":106},2213,"《分层跨模态时空注意力网络用于复杂喜马拉雅农业生态系统作物产量预测》","https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fplant-science\u002Farticles\u002F10.3389\u002Ffpls.2026.1933033\u002Ffull","Mir等提出全面的时空注意力框架,具有四项关键创新:在单树、地块、果园、区域尺度运行的分层注意力机制;融合异构数据流(物联网土壤传感器、气象站、植物生理监测仪、无人机多光谱图像)的跨模态注意力模块;显式建模土壤记忆、滞后和根区耦合的土壤感知注意力头;通过蒙特卡洛Dropout和分位数回归实现不确定性量化。在2023-2026年喜马拉雅苹果园数据集(1247棵监测树)上验证,4周产量预测R²达0.891,RMSE较CNN-LSTM基线降低29.6%,芒果园零样本迁移R²达0.674。",null,"Frontiers in Plant Science 2026年9月10日","2026-09-10T00:00:00Z","论文",10,false,82,{"impact":16,"substance":17,"depth":18,"authority":19,"freshness":20,"relevant":21,"comment":22},18,23,19,14,8,1,"方法新颖、数据规模扎实且含跨作物迁移验证，属智慧农业细分领域高水平研究，值得精选。",[24],{"name":9,"url":6},[26,27,28,29,30,31],"智慧农业","农业人工智能","产量预测","多模态融合","遥感监测","苹果园",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},"H6","biodiversity","生态保护与生物多样性","\u002Ftopic\u002Fenvironment\u002Fbiodiversity","H",[],{"status":40,"locked":13,"version":41,"method":42,"evidence":43},"classified","2026-10-22.2","rules",{"candidates":44,"agricultural_evidence":46},[45,51,59,66,73,79,85],{"code":33,"score":20,"supported":46,"title":47,"abstract":49,"tags":50},true,[48],"农业生态",[],[],{"code":52,"score":53,"supported":46,"title":54,"abstract":55,"tags":58},"B2",6,[],[56,57],"苹果","芒果",[],{"code":60,"score":61,"supported":46,"title":62,"abstract":63,"tags":65},"I3",5,[],[64],"无人机多光谱",[28,30],{"code":67,"score":68,"supported":46,"title":69,"abstract":70,"tags":72},"H2",3,[],[71],"土壤",[],{"code":74,"score":68,"supported":46,"title":75,"abstract":76,"tags":78},"I5",[],[77],"物联网",[],{"code":80,"score":81,"supported":13,"title":82,"abstract":83,"tags":84},"I2",2,[],[],[27,29],{"code":86,"score":21,"supported":13,"title":87,"abstract":88,"tags":89},"I1",[],[],[26],[91,92],"农业人工智能 多模态融合 产量预测 智慧农业","农业人工智能 多模态融合","农业人工智能多模态融合产量预测智慧农业-2213",4,"10.3389\u002Ffpls.2026.1933033\u002Ffull",{"doi":95,"openalex_id":8,"authors":97,"venue":8,"cited_by_count":98,"oa_url":8,"card":99,"direction":103,"ingested_from":105},[],0,{"tldr":100,"method":101,"finding":102,"direction":103,"opportunity":104},"提出分层跨模态时空注意力网络，融合多源数据预测喜马拉雅果园作物产量。","分层注意力+跨模态融合物联网、气象、无人机多光谱数据，蒙特卡洛Dropout量化","4周产量预测R²达0.891，RMSE较CNN-LSTM降低29.6%，芒果园零样本迁移R²达0.6","农业人工智能与决策模型","可探索跨物种零样本迁移的域适应机制，以及轻量化模型在边缘设备上的实时部署。","agent","2026-09-12T00:06:40.371547Z",{"total":53,"page":21,"page_size":53,"items":108},[109,187,263,349,453,502],{"id":110,"title":111,"url":112,"summary":113,"summary_zh":114,"content":8,"source_name":115,"source_url":112,"published_at":116,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":117,"score_detail":118,"sources":123,"tags":125,"primary_topic":130,"secondary_topics":135,"classification":137,"search_phrases":160,"slug":163,"view_count":98,"doi":164,"paper":165,"created_at":186},5392,"Wetland Loss Associated with Hydro-Agricultural Development in North African Floodplains: Insights from a Historical Cartographic Analysis of the Gharb Plain (Morocco)","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fd18100642","Mediterranean floodplain wetlands are among the most productive ecosystems, yet they are increasingly threatened by agricultural intensification and hydrological modifications. In North Africa, most of these wetlands have been converted to agricultural lands using an extensive ‘drainage–irrigation’ process that has irreversibly degraded them. This study documents this loss through the illustrative example of the Gharb Plain in the northwest of Morocco. The agricultural development process of this plain, initiated in 1927 and only recently completed, is documented by both historical maps and remote sensing data, which we combined to reconstruct the simultaneous wetland decline and agriculture extension. The results reveal a dramatic decline in marshland ecosystems, covering about 88% of the original marshlands in less than one cycle. Several wetlands, including large ones, have entirely disappeared, particularly within the Sebou and Baht river floodplains. To ensure their human occupation, the drainage process and river canalization have been supported by a large program of dam construction upstream of the Gharb Plain. The wetland loss estimate highlights the substantial ecological costs of agricultural intensification and underscores the need for integrated management strategies balancing food production and patrimony conservation. Future research should focus on wetland restoration, ecosystem service assessment, and climate-resilient water management in Mediterranean floodplains.","地中海洪泛平原湿地是最具生产力的生态系统之一，却日益受到农业集约化与水文改造的威胁。在北非，大多数此类湿地已通过大规模的“排水—灌溉”过程被转变为农业用地，并因此遭到不可逆的退化。本研究以摩洛哥西北部加尔卜平原（Gharb Plain）为例，记录了这一丧失过程。该平原的农业开发进程始于1927年，直至近期才告完成，我们结合历史地图与遥感数据，重建了湿地衰退与农业扩张同步发生的过程。结果表明，沼泽生态系统急剧减少，在不到一个周期内约88%的原生沼泽地消失。包括大型湿地在内的多处湿地已完全消失，尤其是在塞布河（Sebou）与巴赫特河（Baht）洪泛平原内。为保障人类占用，排水过程与河流渠化得到了加尔卜平原上游大规模水坝建设计划的支持。湿地丧失的估算凸显了农业集约化所带来的巨大生态代价，并强调需要制定兼顾粮食生产与遗产保护的综合管理策略。未来研究应聚焦于湿地恢复、生态系统服务评估以及地中海洪泛平原的气候韧性水资源管理。","Diversity","2026-10-19T00:00:00Z",76,{"impact":119,"substance":120,"depth":16,"authority":121,"freshness":20,"relevant":21,"comment":122},16,21,13,"该研究结合历史地图与遥感数据，量化了北非洪泛平原湿地因农业开发而丧失的情况，方法扎实、结论明确，对农业生态与土地管理有参考价值，但属区域性案例研究，公共影响有限。",[124],{"name":115,"url":112},[126,30,127,128,129],"生态修复","灌溉农业","土地利用变化","湿地保护",{"code":131,"slug":132,"name":133,"path":134,"parent_code":37},"H3","irrigation","水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation",[136],{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},{"status":40,"locked":13,"version":41,"method":42,"evidence":138},{"candidates":139,"agricultural_evidence":46},[140,146,153],{"code":131,"score":12,"supported":46,"title":141,"abstract":142,"tags":145},[],[143,144,132],"灌溉","水资源",[127],{"code":33,"score":147,"supported":46,"title":148,"abstract":149,"tags":152},9,[],[150,151],"生态系统服务","ecosystem service",[128,129,126],{"code":60,"score":154,"supported":46,"title":155,"abstract":156,"tags":159},7,[],[157,158],"遥感","remote sensing",[30],[161,162],"Gharb平原 湿地 遥感","摩洛哥 湿地 灌溉","Gharb平原湿地遥感-5392","10.3390\u002Fd18100642",{"doi":164,"openalex_id":166,"authors":167,"venue":115,"cited_by_count":98,"oa_url":112,"card":179,"direction":183,"ingested_from":185},"W7222966485",[168,171,173,176],{"name":169,"orcid":170},"Mohamed Dakki","https:\u002F\u002Forcid.org\u002F0000-0002-8042-3972",{"name":172,"orcid":8},"Said El Falah",{"name":174,"orcid":175},"Ibrahim Makrane","https:\u002F\u002Forcid.org\u002F0009-0002-9201-206X",{"name":177,"orcid":178},"Ismail Mansouri","https:\u002F\u002Forcid.org\u002F0000-0002-3679-6050",{"tldr":180,"method":181,"finding":182,"direction":183,"opportunity":184},"用历史地图与遥感重建摩洛哥Gharb平原1927年以来的湿地丧失与农业扩张过程。","历史地图与遥感数据结合，重建湿地退缩与农田扩张的时空过程。","不到一个开发周期内约88%的原生沼泽消失，大型湿地完全丧失。","农业遥感与作物表型","可延伸至地中海洪泛平原湿地恢复优先区识别与生态系统服务权衡研究。","openalex","2026-10-20T23:30:19.135602Z",{"id":188,"title":189,"url":190,"summary":191,"summary_zh":192,"content":8,"source_name":193,"source_url":190,"published_at":194,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":195,"score_detail":196,"sources":198,"tags":200,"primary_topic":203,"secondary_topics":204,"classification":211,"search_phrases":236,"slug":239,"view_count":98,"doi":240,"paper":241,"created_at":262},5096,"Land Cover Change, Ecosystem Service Implications, and Economy–Land Use Decoupling:","https:\u002F\u002Fdoi.org\u002F10.14246\u002Firspsd.14.4_186","Land cover change plays a critical role in shaping watershed ecological conditions and supporting regional economic activities. This study analyzes land cover dynamics in the Tiworo Watershed, Southeast Sulawesi, Indonesia, for 2014, 2019, and 2024, and examines their relationship with sectoral economic growth. Remote sensing and GIS were used to map and quantify land cover changes, while sectoral Gross Regional Domestic Product (GRDP) data for agriculture, forestry, and fisheries represented economic performance. Economy–land use interactions were assessed using the Tapio decoupling elasticity index. Economy–land use interactions were assessed using the Tapio decoupling elasticity index as a retrospective empirical measure based on observed changes in land cover and sectoral GRDP over the study periods. The results show increasing sectoral GRDP accompanied by relatively limited changes in dominant land cover classes. Absolute decoupling was observed during 2014–2019, relative decoupling during 2019–2024, and overall absolute decoupling for 2014–2024, indicating that economic growth was not proportionally associated with large-scale land conversion. Spatial analysis reveals localized landscape transformations, including mangrove degradation, settlement expansion, and plantation growth, which signal emerging pressures on environmentally sensitive areas. Although ecosystem services are not directly valued, land cover change is interpreted as an indicator of ecological pressure with implications for provisioning ecosystem services. These findings highlight the importance of integrating land-use monitoring and economic indicators to support adaptive watershed management, spatial planning, and sustainable development.","土地覆盖变化在塑造流域生态状况和支撑区域经济活动方面发挥着关键作用。本研究分析了印度尼西亚东南苏拉威西省蒂沃罗流域（Tiworo Watershed）2014年、2019年和2024年的土地覆盖动态，并考察其与部门经济增长的关系。研究采用遥感和地理信息系统（GIS）对土地覆盖变化进行制图和量化，同时以农业、林业和渔业的部门地区生产总值（GRDP）数据表征经济表现。经济与土地利用的相互作用采用Tapio脱钩弹性指数进行评估，该指数作为基于研究期内观测到的土地覆盖变化和部门GRDP变化的回顾性经验测度。结果表明，部门GRDP持续增长，而主导土地覆盖类型的变化相对有限。2014—2019年呈现绝对脱钩，2019—2024年呈现相对脱钩，2014—2024年整体呈现绝对脱钩，表明经济增长与大规模土地转换之间不存在比例关联。空间分析揭示了局部景观转型，包括红树林退化、居民点扩张和种植园增长，这些信号表明环境敏感区域正面临新兴压力。尽管生态系统服务未被直接估值，但土地覆盖变化被解释为生态压力的指标，并对供给型生态系统服务具有潜在影响。这些发现凸显了整合土地利用监测与经济指标的重要性，以支持适应性流域管理、空间规划和可持续发展。","International Review for Spatial Planning and Sustainable Development","2026-10-14T00:00:00Z",63,{"impact":20,"substance":16,"depth":119,"authority":121,"freshness":20,"relevant":21,"comment":197},"基于遥感与GIS的流域土地利用变化与经济增长脱钩研究，方法规范但属区域案例，对国内三农信息化参考价值有限。",[199],{"name":193,"url":190},[30,128,201,150,202],"流域管理","空间规划",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[205],{"code":206,"slug":207,"name":208,"path":209,"parent_code":210},"I4","gis","GIS与空间分析","\u002Ftopic\u002Fdigital-agriculture\u002Fgis","I",{"status":40,"locked":13,"version":41,"method":42,"evidence":212},{"candidates":213,"agricultural_evidence":46},[214,218,226,230],{"code":33,"score":121,"supported":46,"title":215,"abstract":216,"tags":217},[151],[150],[128,150],{"code":206,"score":12,"supported":46,"title":219,"abstract":220,"tags":225},[],[221,222,223,224],"地理信息","空间分析","GIS","spatial analysis",[202],{"code":60,"score":154,"supported":46,"title":227,"abstract":228,"tags":229},[],[157,158],[30],{"code":231,"score":68,"supported":46,"title":232,"abstract":233,"tags":235},"D",[],[234],"渔业",[],[237,238],"Tiworo流域 土地利用 遥感","Tapio脱钩 农业经济 土地覆盖","Tiworo流域土地利用遥感-5096","10.14246\u002Firspsd.14.4_186",{"doi":240,"openalex_id":242,"authors":243,"venue":193,"cited_by_count":98,"oa_url":256,"card":257,"direction":183,"ingested_from":185},"W7221789653",[244,247,250,253],{"name":245,"orcid":246},"Safril Kasim","https:\u002F\u002Forcid.org\u002F0009-0009-4679-4233",{"name":248,"orcid":249},"Hasddin Hasddin","https:\u002F\u002Forcid.org\u002F0000-0002-0114-5274",{"name":251,"orcid":252},"La Fariki","https:\u002F\u002Forcid.org\u002F0009-0003-7870-9028",{"name":254,"orcid":255},"Abhishek Agarwal","https:\u002F\u002Forcid.org\u002F0000-0002-0900-6383","https:\u002F\u002Fwww.jstage.jst.go.jp\u002Farticle\u002Firspsd\u002F14\u002F4\u002F14_9\u002F_pdf",{"tldr":258,"method":259,"finding":260,"direction":183,"opportunity":261},"分析印尼Tiworo流域2014-2024年土地覆盖变化与部门经济增长的脱钩关系。","遥感与GIS制图，结合GRDP数据和Tapio脱钩弹性指数。","经济持续增长但土地覆盖变化有限，整体呈绝对脱钩，局部红树林退化等生态压力显现。","可将脱钩指数与生态系统服务价值核算结合，量化流域生态压力并支撑空间规划。","2026-10-16T23:30:28.687070Z",{"id":264,"title":265,"url":266,"summary":267,"summary_zh":268,"content":8,"source_name":269,"source_url":266,"published_at":270,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":271,"score_detail":272,"sources":276,"tags":278,"primary_topic":282,"secondary_topics":283,"classification":284,"search_phrases":314,"slug":317,"view_count":21,"doi":318,"paper":319,"created_at":348},4992,"Evaluating tools and methods for agroecosystem sustainability and performance: a critical review","https:\u002F\u002Fdoi.org\u002F10.1007\u002Fs44274-026-01121-4","The sustainability and performance of agroecosystems are critical for confronting global challenges, from food security to biodiversity loss. This critical narrative review synthesizes 137 peer-reviewed sources and grey literature documents (2000-2024) to interrogate the epistemological assumptions, power dynamics, and contextual appropriateness of sustainability assessment frameworks. This comprehensive review systematically analyzes the diverse tools and methodologies; including Life Cycle Assessment (LCA), Response-Inducing Sustainability Evaluation (RISE), Framework for the Evaluation of Natural Resource Management Systems Incorporating Sustainability Indicators (MESMIS), and Tool for Agroecology Performance Evaluation (TAPE); used to evaluate this sustainability. This analysis moves beyond a simple comparison to offer a critical interrogation of these frameworks, exposing a fundamental tension between reductionist tools that offer precise, quantifiable data but often miss systemic interactions, and holistic, participatory approaches that capture socio-ecological complexity but struggle with standardization and scalability. This review underscores that sustainability assessment is not a neutral technical exercise but a value-laden process fraught with trade-offs, power dynamics, and epistemological conflicts, particularly between scientific and indigenous knowledge systems. The transformative potential of emerging technologies like Artificial Intelligence (AI) and remote sensing is examined, while cautioning against technological optimism that risks exacerbating inequities. Ultimately, this review argues for a shift towards reflexive, pluralistic assessment processes that prioritize equity, adaptive learning, and critical independence over standardized scoring, providing practical heuristics for researchers and practitioners navigating methodological trade-offs in diverse contexts.","农业生态系统的可持续性与表现对于应对从粮食安全到生物多样性丧失等全球性挑战至关重要。本批判性叙事综述综合了137篇同行评审文献和灰色文献（2000—2024年），旨在审视可持续性评估框架的认识论假设、权力动态及情境适宜性。本综述系统分析了用于评估该可持续性的多种工具和方法，包括生命周期评价（LCA）、诱导响应可持续性评估（RISE）、纳入可持续性指标的自然资源管理系统评价框架（MESMIS）以及农业生态表现评估工具（TAPE）。本分析超越了简单的比较，对这些框架进行了批判性审视，揭示了还原论工具与整体性参与式方法之间的根本张力——前者提供精确、可量化的数据，但往往忽视系统性交互；后者捕捉社会生态复杂性，却在标准化和可扩展性方面面临困难。本综述强调，可持续性评估并非中性的技术操作，而是一个充满价值负载的过程，充斥着权衡取舍、权力动态和认识论冲突，尤其是科学知识体系与本土知识体系之间的冲突。本文审视了人工智能（AI）和遥感等新兴技术的变革潜力，同时警示技术乐观主义可能加剧不平等。最终，本综述主张转向反思性、多元化的评估过程，将公平、适应性学习和批判性独立置于标准化评分之上，并为研究者和实践者在不同情境中应对方法论权衡提供实用启发式指南。","Discover Environment","2026-10-13T00:00:00Z",71,{"impact":273,"substance":274,"depth":16,"authority":121,"freshness":20,"relevant":21,"comment":275},12,20,"系统梳理LCA、RISE、MESMIS、TAPE等农业生态可持续性评估工具，指出还原论与整体论方法的张力，并讨论AI与遥感的赋能潜力，对农业信息化与智慧农业评估体系有参考价值。",[277],{"name":269,"url":266},[27,279,48,280,281],"农业遥感","生命周期评价","可持续评价",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":13,"version":41,"method":42,"evidence":285},{"candidates":286,"agricultural_evidence":46},[287,293,297,304,309],{"code":33,"score":12,"supported":46,"title":288,"abstract":289,"tags":292},[],[290,48,34,291],"生物多样性","agroecology",[48],{"code":60,"score":154,"supported":46,"title":294,"abstract":295,"tags":296},[],[157,158],[279],{"code":298,"score":53,"supported":46,"title":299,"abstract":300,"tags":303},"A8",[],[301,302],"粮食安全","food security",[],{"code":80,"score":94,"supported":46,"title":305,"abstract":306,"tags":308},[],[307],"artificial intelligence",[27],{"code":310,"score":21,"supported":13,"title":311,"abstract":312,"tags":313},"H4",[],[],[280],[315,316],"TAPE 农业生态 评价","RISE MESMIS 可持续评估","TAPE农业生态评价-4992","10.1007\u002Fs44274-026-01121-4",{"doi":318,"openalex_id":320,"authors":321,"venue":269,"cited_by_count":98,"oa_url":340,"card":341,"direction":347,"ingested_from":185},"W7221633095",[322,325,328,331,333,335,337],{"name":323,"orcid":324},"Pierre Marie Chimi","https:\u002F\u002Forcid.org\u002F0000-0003-4520-4165",{"name":326,"orcid":327},"Karimou Ngamsou Abdel","https:\u002F\u002Forcid.org\u002F0000-0002-8126-6235",{"name":329,"orcid":330},"Baruch Batamack Nkoué","https:\u002F\u002Forcid.org\u002F0000-0001-5601-3327",{"name":332,"orcid":8},"Eusebe Ydelphonse Nyonce Pokam",{"name":334,"orcid":8},"Angele Marie Ngon Dikoume",{"name":336,"orcid":8},"Benjamin Loïc Ndong Ndzana",{"name":338,"orcid":339},"William Armand Mala","https:\u002F\u002Forcid.org\u002F0000-0002-2240-3746","https:\u002F\u002Flink.springer.com\u002Fcontent\u002Fpdf\u002F10.1007\u002Fs44274-026-01121-4.pdf",{"tldr":342,"method":343,"finding":344,"direction":345,"opportunity":346},"综述137篇文献，批判性评估农业生态系统可持续性评价工具的方法论假设与权衡。","批判性叙事综述，分析LCA、RISE、MESMIS、TAPE等框架及AI与遥感。","评价工具在简化精确与整体参与间存在根本张力，评价是价值负载过程。","农业绿色发展与碳","可开发兼顾标准化与参与式、融合本土知识并反思技术乐观的混合评价框架。","智慧农业 \u002F 农业物联网","2026-10-15T23:30:10.405237Z",{"id":350,"title":351,"url":352,"summary":353,"summary_zh":354,"content":8,"source_name":355,"source_url":352,"published_at":194,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":356,"score_detail":357,"sources":360,"tags":362,"primary_topic":363,"secondary_topics":367,"classification":377,"search_phrases":426,"slug":429,"view_count":21,"doi":430,"paper":431,"created_at":452},4988,"Hierarchical cross-modal spatiotemporal attention network for crop yield prediction in complex Himalayan agroecosystems","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffpls.2026.1933033","Introduction Accurate crop yield prediction is essential for precision agriculture, climate-resilient farm management, and timely decision-making. However, yield prediction remains challenging because of complex, nonlinear, and multi-scale interactions among crop physiology, soil conditions, weather variability, and management practices. Conventional statistical, machine learning, and deep learning approaches, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory (LSTM) networks, often have limitations in capturing long-range temporal dependencies, hierarchical spatial relationships, heterogeneous data modalities, and predictive uncertainty. Methods This study proposes a Hierarchical Cross-Modal Spatio-Temporal Attention Network (HCM-STAN) for crop yield prediction using a unified attention-based framework for integrating heterogeneous agro-environmental data. The architecture incorporates four principal components: (i) hierarchical spatio-temporal attention operating across tree, block, orchard, and regional scales; (ii) a cross-modal attention module for adaptive fusion of IoT-based soil sensor observations, weather data, plant physiological measurements, and drone-based multispectral imagery; (iii) a soil-aware attention head that explicitly models soil memory, hysteresis effects, and root-zone dynamics; and (iv) an uncertainty-aware prediction module combining Monte Carlo dropout and quantile regression to generate calibrated prediction intervals. The framework was evaluated using a multi-source dataset collected from Himalayan apple orchards during 2023–2026, comprising 1,247 monitored trees across three commercial orchards. Results The proposed HCM-STAN achieved an R 2 of 0.89 for four-week-ahead crop yield prediction, representing a 15.6% improvement over the CNN-LSTM baseline. Furthermore, zero-shot transfer to mango orchards achieved an R 2 of 0.67, corresponding to approximately 75% of the performance obtained under full training. Discussion The results demonstrate that integrating hierarchical spatio-temporal representation learning, cross-modal attention, soil-aware modelling, and uncertainty estimation can effectively address the heterogeneous and dynamic nature of crop yield prediction in complex mountain agroecosystems. The observed zero-shot transfer performance further indicates the potential of HCM-STAN to generalize across crop systems without crop-specific retraining. Overall, the proposed framework provides a promising foundation for reliable, uncertainty-aware yield forecasting and can support precision agriculture, climate-smart farm management, and data-driven decision-making in heterogeneous agroecosystems.","引言 准确的作物产量预测对于精准农业、气候适应性农场管理以及及时决策至关重要。然而，由于作物生理、土壤条件、天气变异和管理措施之间复杂、非线性且多尺度的相互作用，产量预测仍然具有挑战性。传统的统计方法、机器学习以及深度学习方法，包括卷积神经网络（CNN）、循环神经网络（RNN）和长短期记忆（LSTM）网络，在捕捉长程时间依赖关系、层次化空间关系、异构数据模态以及预测不确定性方面往往存在局限。方法 本研究提出了一种用于作物产量预测的层次化跨模态时空注意力网络（HCM-STAN），采用统一的基于注意力的框架来整合异构农业环境数据。该架构包含四个主要组成部分：（i）在树、地块、果园和区域尺度上运行的层次化时空注意力；（ii）用于自适应融合基于物联网的土壤传感器观测、天气数据、植物生理测量和无人机多光谱影像的跨模态注意力模块；（iii）显式建模土壤记忆、滞后效应和根区动态的土壤感知注意力头；（iv）结合蒙特卡洛 dropout 和分位数回归以生成校准预测区间的不确定性感知预测模块。该框架使用2023—2026年在喜马拉雅苹果园采集的多源数据集进行了评估，该数据集包含三个商业果园中1，247棵监测树木。结果 所提出的 HCM-STAN 在提前四周作物产量预测中达到了0.89的 R²，相较于 CNN-LSTM 基线提升了15.6%。此外，零样本迁移至芒果园达到了0.67的 R²，相当于完整训练下所获性能的约75%。讨论 结果表明，整合层次化时空表征学习、跨模态注意力、土壤感知建模和不确定性估计，能够有效应对复杂山地农业生态系统中作物产量预测的异构性和动态性。观察到的零样本迁移性能进一步表明，HCM-STAN 具有在无需作物特定再训练的情况下跨作物系统泛化的潜力。总体而言，该 p","Frontiers in Plant Science",81,{"impact":16,"substance":358,"depth":16,"authority":19,"freshness":147,"relevant":21,"comment":359},22,"提出分层跨模态时空注意力网络，融合物联网土壤传感、气象、生理与无人机多光谱数据，苹果园四周产量预测R²达0.89并实现芒果园零样本迁移，方法新颖、数据扎实，对智慧农业产量预报有参考价值。",[361],{"name":355,"url":352},[26,27,28,29,157],{"code":80,"slug":364,"name":365,"path":366,"parent_code":210},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai",[368,373],{"code":52,"slug":369,"name":370,"path":371,"parent_code":372},"fruit","果树","\u002Ftopic\u002Fhorticulture\u002Ffruit","B",{"code":60,"slug":374,"name":375,"path":376,"parent_code":210},"remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing",{"status":40,"locked":13,"version":41,"method":42,"evidence":378},{"candidates":379,"agricultural_evidence":46},[380,389,395,400,406,411,416,420],{"code":80,"score":381,"supported":46,"title":382,"abstract":383,"tags":388},11,[],[384,385,386,387],"深度学习","机器学习","machine learning","deep learning",[27,29],{"code":52,"score":147,"supported":46,"title":390,"abstract":391,"tags":394},[],[56,57,392,393],"orchard","apple",[],{"code":60,"score":20,"supported":46,"title":396,"abstract":397,"tags":399},[],[64,398],"multispectral",[28,157],{"code":86,"score":154,"supported":46,"title":401,"abstract":402,"tags":405},[],[403,404],"精准农业","precision agriculture",[26],{"code":67,"score":53,"supported":46,"title":407,"abstract":408,"tags":410},[],[71,409],"soil",[],{"code":74,"score":53,"supported":46,"title":412,"abstract":413,"tags":415},[],[77,414],"IoT",[],{"code":33,"score":68,"supported":46,"title":417,"abstract":418,"tags":419},[],[48],[],{"code":421,"score":68,"supported":46,"title":422,"abstract":423,"tags":425},"H7",[],[424],"气候适应",[],[427,428],"喜马拉雅 苹果园 产量预测","HCM-STAN 作物产量预测","喜马拉雅苹果园产量预测-4988","10.3389\u002Ffpls.2026.1933033",{"doi":430,"openalex_id":432,"authors":433,"venue":355,"cited_by_count":98,"oa_url":352,"card":447,"direction":347,"ingested_from":185},"W7221795850",[434,437,439,441,443,445],{"name":435,"orcid":436},"Shabir Ahmad Mir","https:\u002F\u002Forcid.org\u002F0000-0003-2311-218X",{"name":438,"orcid":8},"Umarfarooque Momin",{"name":440,"orcid":8},"Sudeep Kumar E",{"name":442,"orcid":8},"Itfaq Ahmad Mir",{"name":444,"orcid":8},"Ab Rouf Khan",{"name":446,"orcid":8},"Ghulam Mohammad Mir",{"tldr":448,"method":449,"finding":450,"direction":103,"opportunity":451},"提出分层跨模态时空注意力网络HCM-STAN，融合多源农业数据预测喜马拉雅山区苹果产量。","用IoT土壤、气象、生理和无人机多光谱数据，结合分层时空与跨模态注意力及不确定性","四周前产量预测R²达0.89，比CNN-LSTM提升15.6%，零样本迁移芒果R²为0.67。","可探索跨作物零样本迁移的领域自适应机制，并验证模型在更多山地农业生态系统的泛化能力。","2026-10-15T23:30:08.750140Z",{"id":454,"title":455,"url":456,"summary":457,"summary_zh":458,"content":8,"source_name":459,"source_url":456,"published_at":460,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":356,"score_detail":461,"sources":463,"tags":465,"primary_topic":469,"secondary_topics":470,"classification":471,"search_phrases":482,"slug":485,"view_count":98,"doi":486,"paper":487,"created_at":501},4821,"Producing a Red List of Ecosystems for the Brazilian Cerrado","https:\u002F\u002Fdoi.org\u002F10.1111\u002Fcsp2.70426","Abstract Land clearance is a key driver of biodiversity loss. This is a particular challenge in the Brazilian Cerrado, the world's most biodiverse savanna biome and a region experiencing a rapid expansion of agriculture. To inform habitat conservation and restoration actions, we used the Red List of Ecosystems (RLE) framework to identify threatened ecosystems within the Cerrado. This biome is divided into 20 ecoregions, and we subdivided each ecoregion into forest, savanna, grassland, and wetland ecosystems. We used remote sensing data from 1985 to 2020 to measure and predict the distribution of each ecosystem and assess its risk of collapse based on Criterion A2b (reduction in geographic distribution over any 50‐year period), Criterion B1 (limited extent of occurrence), and Criterion B2 (limited area of occupancy). Of the 80 ecosystems comprising the Cerrado, we assessed 12 as Endangered and 43 as Vulnerable. Most of these threatened ecosystems were identified under Criterion A2b, while Criterion B identified a smaller subset of spatially restricted ecosystems, particularly wetlands, underscoring the importance of applying more than one criterion in the assessment process. By using the Red List of Ecosystems as an international standard, we provide the first globally comparable analysis of the state of the Cerrado, showing that nearly three‐quarters of the region's ecosystems are at risk of collapse. This information provides a baseline to inform local policy and practice, helping to conserve biodiversity and maintain ecosystem services within this globally important region.","土地开垦是生物多样性丧失的主要驱动因素。这在巴西塞拉多（Cerrado）尤为严峻，该地区拥有全球生物多样性最丰富的稀树草原生物群系，同时正经历农业的快速扩张。为了为栖息地保护和恢复行动提供依据，我们采用生态系统红色名录（Red List of Ecosystems, RLE）框架来识别塞拉多境内受威胁的生态系统。该生物群系被划分为20个生态区，我们又将每个生态区细分为森林、稀树草原、草地和湿地生态系统。我们利用1985年至2020年的遥感数据，测量并预测各生态系统的分布，并基于标准A2b（任意50年期内地理分布范围的缩减）、标准B1（分布范围有限）和标准B2（占据面积有限）评估其崩溃风险。在构成塞拉多的80个生态系统中，我们评估出12个为濒危，43个为易危。这些受威胁的生态系统大多依据标准A2b被识别，而标准B则识别出较少一部分空间受限的生态系统，尤其是湿地，这凸显了在评估过程中应用多项标准的重要性。通过将生态系统红色名录作为国际标准加以运用，我们提供了首个在全球范围内可比较的塞拉多状况分析，表明该地区近四分之三的生态系统面临崩溃风险。这些信息为地方政策和实践提供了基线，有助于在这一具有全球重要性的区域内保护生物多样性并维持生态系统服务。","Conservation Science and Practice","2026-10-12T00:00:00Z",{"impact":16,"substance":358,"depth":16,"authority":19,"freshness":147,"relevant":21,"comment":462},"基于1985—2020年遥感数据对巴西塞拉多80个生态系统做红色名录评估，方法规范、数据扎实，对农业扩张与生态保护平衡有参考价值，但属域外研究，国内落地影响有限。",[464],{"name":459,"url":456},[290,30,466,467,468],"农业扩张","草原保护","生态系统评估",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":13,"version":41,"method":42,"evidence":472},{"candidates":473,"agricultural_evidence":46},[474,478],{"code":33,"score":12,"supported":46,"title":475,"abstract":476,"tags":477},[],[290,150,34],[290],{"code":60,"score":154,"supported":46,"title":479,"abstract":480,"tags":481},[],[157,158],[30],[483,484],"巴西塞拉多 生态系统红色名录","塞拉多 遥感 生物多样性","巴西塞拉多生态系统红色名录-4821","10.1111\u002Fcsp2.70426",{"doi":486,"openalex_id":488,"authors":489,"venue":459,"cited_by_count":98,"oa_url":456,"card":496,"direction":183,"ingested_from":185},"W7221517508",[490,493],{"name":491,"orcid":492},"Juliana Ávila Martins","https:\u002F\u002Forcid.org\u002F0000-0002-5465-8276",{"name":494,"orcid":495},"Robert J. Smith","https:\u002F\u002Forcid.org\u002F0000-0003-1599-9171",{"tldr":497,"method":498,"finding":499,"direction":183,"opportunity":500},"用生态系统红色名录框架评估巴西塞拉多80个生态系统的崩溃风险。","基于1985-2020年遥感数据，按A2b、B1、B2标准评估分布与风险。","近四分之三生态系统受威胁，12个濒危、43个易危，湿地等空间受限类型突出。","可结合多源遥感与生态模型，动态监测农业扩张下生态系统风险并支撑保护修复决策。","2026-10-13T23:30:23.689100Z",{"id":503,"title":504,"url":505,"summary":506,"summary_zh":507,"content":8,"source_name":508,"source_url":505,"published_at":509,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":510,"score_detail":511,"sources":513,"tags":515,"primary_topic":517,"secondary_topics":518,"classification":519,"search_phrases":530,"slug":533,"view_count":21,"doi":534,"paper":535,"created_at":550},4738,"Natural ecosystem loss in an Atlantic Forest agricultural basin: drivers of land-use change and future landscape trajectories","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffsufs.2026.1912707","The Atlantic Forest, one of the world's most threatened biodiversity hotspots, faces intense pressure from agricultural expansion and inadequate land governance. Understanding the spatial drivers and future trajectories of land-use and land cover (LULC) change is critical for safeguarding the ecosystem services and biodiversity that remnant natural areas provide. This study investigated the relative importance of multiple biophysical, socioeconomic, and demographic drivers of natural ecosystem loss in the Guaporé River Basin (GRB), a 246,300-ha subtropical Atlantic Forest agricultural watershed in southern Brazil. We combined a municipality-level temporal trend analysis to characterize historical LULC dynamics (1998–2019) with the Dinamica EGO cellular automata model, calibrated using Weight of Evidence (WoE), to simulate future landscapes under 2030 and 2050 scenarios. Our results reveal divergent ecological dynamics across the basin. The Northern region experienced extensive remnant vegetation loss (1.55% yr −1 conversion to altered ecosystems), driven by flat terrain and high agricultural productivity, while the Southern region showed partial natural regeneration (1.40% yr −1 transition from altered to regenerating vegetation), linked to steeper slopes and vegetation extraction practices. Despite apparent recovery, closer analysis reveals a regressive process of ecological simplification, in which remnant vegetation is progressively replaced by regenerating vegetation. Future projections suggest that, without strengthened enforcement of Brazil's Natural Vegetation Protection Law, the Northern basin will lose an additional 15,903 ha of remnant vegetation by 2050. These findings highlight critical spatial mismatches between legal protection frameworks and actual land-use pressures, with direct implications for conservation planning, watershed management, and biodiversity protection in the Atlantic Forest biome.","大西洋森林是全球受威胁最严重的生物多样性热点地区之一，面临着农业扩张和土地治理不足带来的巨大压力。理解土地利用与土地覆盖（LULC）变化的空间驱动因素及未来演变趋势，对于维护残余自然区域所提供的生态系统服务和生物多样性至关重要。本研究考察了巴西南部瓜波雷河流域（GRB）自然生态系统丧失的多个生物物理、社会经济和人口驱动因素的相对重要性。该流域面积246,300公顷，属亚热带大西洋森林农业流域。我们将市级时间趋势分析（用于刻画1998—2019年历史LULC动态）与Dinamica EGO元胞自动机模型相结合，并采用证据权重法（WoE）进行校准，以模拟2030年和2050年情景下的未来景观。研究结果揭示了流域内差异化的生态动态。北部地区经历了大面积的残余植被丧失（年转化率1.55%，转为改变后的生态系统），其驱动因素为平坦地形和高农业生产力；而南部地区则表现出部分自然再生（年转化率1.40%，从改变后的生态系统转为再生植被），与较陡的坡度和植被采伐活动相关。尽管表面上呈现恢复态势，进一步分析却揭示了一个退行性的生态简化过程，即残余植被正逐步被再生植被所取代。未来预测表明，若不加紧执行巴西《自然植被保护法》，到2050年北部流域将再丧失15,903公顷残余植被。这些发现凸显了法律保护框架与实际土地利用压力之间的关键空间错配，对大西洋森林生物群系的保护规划、流域管理和生物多样性保护具有直接影响。","Frontiers in Sustainable Food Systems","2026-10-11T00:00:00Z",75,{"impact":273,"substance":358,"depth":16,"authority":19,"freshness":147,"relevant":21,"comment":512},"基于元胞自动机与证据权重的巴西大西洋森林流域土地利用模拟研究，方法扎实、结论具体，但属区域案例，公共影响有限。",[514],{"name":508,"url":505},[48,290,30,128,516],"流域治理",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":13,"version":41,"method":42,"evidence":520},{"candidates":521,"agricultural_evidence":46},[522,526],{"code":33,"score":273,"supported":46,"title":523,"abstract":524,"tags":525},[],[290,150,34],[48,128,290],{"code":60,"score":21,"supported":13,"title":527,"abstract":528,"tags":529},[],[],[30],[531,532],"Guaporé River Basin 土地利用","Atlantic Forest 农业流域 模拟","GuaporéRiverBasin土地利用-4738","10.3389\u002Ffsufs.2026.1912707",{"doi":534,"openalex_id":536,"authors":537,"venue":508,"cited_by_count":98,"oa_url":505,"card":545,"direction":183,"ingested_from":185},"W7221477243",[538,541,543],{"name":539,"orcid":540},"Edberto Moura Lima","https:\u002F\u002Forcid.org\u002F0000-0002-8447-8460",{"name":542,"orcid":8},"Bano Mehdi-Schulz",{"name":544,"orcid":8},"Jose Miguel Reichert",{"tldr":546,"method":547,"finding":548,"direction":183,"opportunity":549},"结合趋势分析与元胞自动机模拟巴西大西洋森林农业流域土地利用变化及未来情景。","Dinamica EGO元胞自动机+证据权重法，1998–2019年市级LULC","北部残存植被持续丧失，南部看似恢复实为生态简化，2050年北部将再失1.59万公顷。","可引入遥感时序与多尺度治理数据，量化法律保护与实际压力错配的驱动机制。","2026-10-12T23:30:06.576251Z"]