[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3516":3,"related-3516":124},{"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":23,"tags":25,"primary_topic":31,"secondary_topics":37,"classification":44,"search_phrases":91,"slug":94,"view_count":95,"doi":96,"paper":97,"created_at":123},3516,"Approaches to forecast soil nutrient dynamics for precision agriculture and sustainable fertiliser management: A review","https:\u002F\u002Fdoi.org\u002F10.14719\u002Fpst.16160","Predictive modelling of soil nutrient dynamics is an essential tool for promoting sustainable agricultural practices and environmentally responsible farming methods. The statistical and machine learning techniques used to forecast the availability and dynamics of soil nutrients are summarised in this review. The core frameworks for measuring spatio-temporal nutritional variability are established by traditional statistical approaches such as time-series models autoregressive integrated moving average (ARIMA), seasonal autoregressive integrated moving average (SARIMA), multivariate techniques (Principal component analysis (PCA) and factor analysis) and geostatistical tools (kriging). By capturing intricate nonlinear interactions within heterogeneous agroecosystems, machine learning techniques like random forest, support vector machines and ensemble approaches (XGBoost, LightGBM and AdaBoost) provide higher prediction accuracy. Forecasting capabilities are further enhanced by hybrid frameworks [Autoregressive integrated moving average with exogenous variables–artificial neural network. (ARIMAX-ANN)] and deep learning architectures (Convolutional neural network (CNN), long short-term memory (LSTM), ANN). With R2 values above 0.93 and notable decreases in prediction errors, ensemble approaches routinely perform better than traditional linear models. Nevertheless, persistent challenges include data quality limitations, spatial sampling constraints, insufficient environmental covariates and reduced model transferability across diverse pedoclimatic regions. Integrating high-resolution soil properties, climatic variables, terrain attributes and spectral information with advanced modelling architectures remains crucial for enhancing predictive reliability, ultimately supporting precision nutrient management, improved fertiliser efficiency and environmentally responsible agricultural systems.","土壤养分动态的预测建模是推动可持续农业实践和环境友好型耕作方法的重要工具。本综述总结了用于预测土壤养分有效性及其动态变化的统计与机器学习技术。传统统计方法，如时间序列模型自回归积分滑动平均模型（ARIMA）、季节性自回归积分滑动平均模型（SARIMA）、多变量技术（主成分分析（PCA）和因子分析）以及地统计工具（克里金法），为量化养分的时空变异性奠定了核心框架。通过捕捉异质性农业生态系统中复杂的非线性相互作用，随机森林、支持向量机和集成方法（XGBoost、LightGBM和AdaBoost）等机器学习技术可实现更高的预测精度。混合框架[含外生变量的自回归积分滑动平均模型–人工神经网络（ARIMAX-ANN）]和深度学习架构（卷积神经网络（CNN）、长短期记忆网络（LSTM）、人工神经网络（ANN））进一步增强了预测能力。集成方法的R²值超过0.93，且预测误差显著降低，其表现通常优于传统线性模型。然而，持续存在的挑战包括数据质量限制、空间采样约束、环境协变量不足以及模型在不同土壤气候区域间可迁移性降低等问题。将高分辨率土壤属性、气候变量、地形属性和光谱信息与先进建模架构相结合，对于提高预测可靠性仍然至关重要，最终可为精准养分管理、提高肥料利用效率以及环境友好型农业系统提供支撑。",null,"Plant Science Today","2026-09-24T00:00:00Z","论文",10,false,79,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,13,8,1,"系统综述土壤养分动态预测的统计与机器学习方法，方法体系完整、结论有量化支撑，对精准施肥与农业信息化有参考价值，但属综述类论文，产业级影响有限。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","变量施肥","机器学习","精准农业","土壤养分",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},"H2","soil","土壤健康与改良","\u002Ftopic\u002Fenvironment\u002Fsoil","H",[38],{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",{"status":45,"locked":14,"version":46,"method":47,"evidence":48},"classified","2026-10-22.2","rules",{"candidates":49,"agricultural_evidence":52},[50,57,63,70,78,85],{"code":32,"score":51,"supported":52,"title":53,"abstract":54,"tags":56},15,true,[33],[55,33],"土壤",[30],{"code":58,"score":13,"supported":52,"title":59,"abstract":61,"tags":62},"I1",[60],"precision agriculture",[],[26,29],{"code":39,"score":13,"supported":52,"title":64,"abstract":65,"tags":69},[],[66,28,67,68],"深度学习","machine learning","deep learning",[28],{"code":71,"score":72,"supported":52,"title":73,"abstract":74,"tags":77},"F3",7,[],[75,76],"肥料","养分管理",[27],{"code":79,"score":80,"supported":52,"title":81,"abstract":82,"tags":84},"H6",3,[],[83],"农业生态",[],{"code":86,"score":80,"supported":52,"title":87,"abstract":88,"tags":90},"H8",[],[89],"可持续农业",[],[92,93],"土壤养分 预测模型 精准农业","机器学习 施肥管理 可持续农业","土壤养分预测模型精准农业-3516",4,"10.14719\u002Fpst.16160",{"doi":96,"openalex_id":98,"authors":99,"venue":10,"cited_by_count":115,"oa_url":6,"card":116,"direction":120,"ingested_from":122},"W7214167059",[100,103,106,109,112],{"name":101,"orcid":102},"R Rathna","https:\u002F\u002Forcid.org\u002F0009-0004-7797-2673",{"name":104,"orcid":105},"B Sivasankari","https:\u002F\u002Forcid.org\u002F0000-0001-9921-8170",{"name":107,"orcid":108},"R. Gangai Selvi","https:\u002F\u002Forcid.org\u002F0000-0002-4475-2293",{"name":110,"orcid":111},"J Prabhakaran","https:\u002F\u002Forcid.org\u002F0000-0001-7339-175X",{"name":113,"orcid":114},"K. G. Sabarinathan","https:\u002F\u002Forcid.org\u002F0000-0002-8659-6479",0,{"tldr":117,"method":118,"finding":119,"direction":120,"opportunity":121},"综述土壤养分动态预测的统计与机器学习方法，比较精度与局限。","综述ARIMA、地统计、随机森林、XGBoost、CNN\u002FLSTM及混合模型。","集成与深度学习模型精度更高（R²>0.93），但数据质量与跨区迁移性仍是瓶颈。","农业人工智能与决策模型","可研究多源遥感与气候数据融合的迁移学习模型，提升跨区域养分预测泛化能力。","openalex","2026-09-25T23:30:54.950445Z",{"total":125,"page":21,"page_size":125,"items":126},6,[127,217,280,331,405,478],{"id":128,"title":129,"url":130,"summary":131,"summary_zh":132,"content":9,"source_name":133,"source_url":130,"published_at":134,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":135,"score_detail":136,"sources":141,"tags":143,"primary_topic":147,"secondary_topics":153,"classification":156,"search_phrases":192,"slug":195,"view_count":115,"doi":196,"paper":197,"created_at":216},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":137,"substance":138,"depth":139,"authority":19,"freshness":13,"relevant":21,"comment":140},16,21,17,"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。",[142],{"name":133,"url":130},[26,28,144,145,146],"遥感监测","土壤墒情","冬小麦",{"code":148,"slug":149,"name":150,"path":151,"parent_code":152},"A2","wheat","小麦","\u002Ftopic\u002Ffield-crops\u002Fwheat","A",[154,155],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},{"status":45,"locked":14,"version":46,"method":47,"evidence":157},{"candidates":158,"agricultural_evidence":52},[159,165,169,173,181,188],{"code":148,"score":160,"supported":52,"title":161,"abstract":163,"tags":164},26,[162,149],"winter wheat",[150,162,149],[146],{"code":32,"score":51,"supported":52,"title":166,"abstract":167,"tags":168},[33],[55,33],[145],{"code":39,"score":51,"supported":52,"title":170,"abstract":171,"tags":172},[67],[28,67],[28],{"code":174,"score":13,"supported":52,"title":175,"abstract":176,"tags":180},"I3",[],[177,178,179],"遥感","remote sensing","multispectral",[144],{"code":182,"score":125,"supported":52,"title":183,"abstract":184,"tags":187},"H3",[],[185,186],"灌溉","irrigation",[],{"code":58,"score":95,"supported":52,"title":189,"abstract":190,"tags":191},[],[26],[26],[193,194],"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":196,"openalex_id":198,"authors":199,"venue":133,"cited_by_count":115,"oa_url":130,"card":209,"direction":215,"ingested_from":122},"W7223156204",[200,203,205,207],{"name":201,"orcid":202},"Mingliang Jiang","https:\u002F\u002Forcid.org\u002F0000-0002-6486-6087",{"name":204,"orcid":9},"Yi Zhang",{"name":206,"orcid":9},"WEI Lei",{"name":208,"orcid":9},"Chengcai Zhang",{"tldr":210,"method":211,"finding":212,"direction":213,"opportunity":214},"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。","农业遥感与作物表型","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":218,"title":219,"url":220,"summary":221,"summary_zh":222,"content":9,"source_name":223,"source_url":220,"published_at":224,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":225,"score_detail":226,"sources":230,"tags":232,"primary_topic":236,"secondary_topics":237,"classification":239,"search_phrases":260,"slug":263,"view_count":80,"doi":264,"paper":265,"created_at":279},4374,"AI-Driven Crop Recommendation and Growth Prediction System Using Soil Fertility and Climate Intelligence","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23179023","Agricultural decisions are increasingly affected by changing climate conditions, soil fertility, and the availability of timely technical guidance. This mini project presents an AI-Driven Agricultural Intelligence System that combines crop recommendation, soil fertility assessment, weather-risk prediction, yield estimation, fertilizer recommendation, and multilingual conversational support in one web platform. The system accepts soil and climate parameters such as nitrogen, phosphorus, potassium, pH, temperature, humidity, rainfall, location, crop, and season. The supplied implementation uses a Random Forest model for crop recommendation, XGBoost for weather-risk prediction, a Random Forest regressor for yield estimation, a Random Forest classifier for fertilizer recommendation, and a deterministic rule-based soilfertility module with model-independent fallback logic. A FastAPI backend performs validation, model loading, chained inference, weather integration, authentication, and response formatting, while a React-based Progressive Web Application provides the user interface. OpenWeatherMap supplies current weather context and Google Gemini AI supports multilingual agricultural explanations with offline FAQ and rulebased fallbacks. Reported module performance is approximately 88-95%, and representative API, fallback, authentication, translation, rate-limiting, and chatbot tests pass in the supplied project material. The project demonstrates how machine learning, web development, external APIs, and multilingual AI can be combined to deliver practical, accessible, and data-driven agricultural decision support. Keywords: Machine Learning, Agriculture, Crop Recommendation, Soil Fertility, Weather Risk, Yield Prediction, Fertilizer Recommendation, FastAPI, React, XGBoost, Google Gemini AI, Precision Agriculture.","农业决策日益受到气候条件变化、土壤肥力以及及时技术指导可得性的影响。本小型项目提出了一套人工智能驱动的农业智能系统，将作物推荐、土壤肥力评估、天气风险预测、产量估算、肥料推荐和多语言对话支持集成于一个网络平台。该系统接受土壤和气候参数，如氮、磷、钾、pH值、温度、湿度、降雨量、地点、作物和季节。所提供的实现采用随机森林（Random Forest）模型进行作物推荐，采用XGBoost进行天气风险预测，采用随机森林回归器进行产量估算，采用随机森林分类器进行肥料推荐，并采用基于确定性规则的土壤肥力模块，具备与模型无关的回退逻辑。FastAPI后端执行验证、模型加载、链式推理、天气集成、身份认证和响应格式化，而基于React的渐进式网络应用提供用户界面。OpenWeatherMap提供当前天气上下文，Google Gemini AI支持多语言农业解释，并具备离线FAQ和基于规则的回退机制。所报告的模块性能约为88-95%，且在所提供的项目材料中，具有代表性的API、回退、身份认证、翻译、速率限制和聊天机器人测试均通过。该项目展示了如何将机器学习、网络开发、外部API和多语言AI相结合，以提供实用、可及且数据驱动的农业决策支持。关键词：机器学习、农业、作物推荐、土壤肥力、天气风险、产量预测、肥料推荐、FastAPI、React、XGBoost、Google Gemini AI、精准农业。","Zenodo (CERN European Organization for Nuclear Research)","2026-10-06T00:00:00Z",50,{"impact":20,"substance":227,"depth":19,"authority":125,"freshness":228,"relevant":21,"comment":229},14,9,"学生级迷你项目论文，方法组合常规、性能数据自报，工程整合有一定参考价值但学术与产业增量有限，可作智慧农业应用案例收录。",[231],{"name":223,"url":220},[26,233,29,234,235],"农业人工智能","作物推荐","土壤肥力",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[238],{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},{"status":45,"locked":14,"version":46,"method":47,"evidence":240},{"candidates":241,"agricultural_evidence":52},[242,246,251,255],{"code":32,"score":51,"supported":52,"title":243,"abstract":244,"tags":245},[33],[55,33],[235],{"code":39,"score":13,"supported":52,"title":247,"abstract":248,"tags":250},[],[28,67,249],"decision support",[233],{"code":58,"score":20,"supported":52,"title":252,"abstract":253,"tags":254},[],[29,60],[26,29],{"code":71,"score":125,"supported":52,"title":256,"abstract":257,"tags":259},[],[75,258],"fertilizer",[],[261,262],"AI 作物推荐 土壤肥力","XGBoost 天气风险 产量预测","AI作物推荐土壤肥力-4374","10.5281\u002Fzenodo.23179023",{"doi":264,"openalex_id":266,"authors":267,"venue":223,"cited_by_count":115,"oa_url":220,"card":274,"direction":120,"ingested_from":122},"W7220664944",[268,270,272],{"name":269,"orcid":9},"Gunashekar B.",{"name":271,"orcid":9},"Saiteja G.",{"name":273,"orcid":9},"Maniteja K.",{"tldr":275,"method":276,"finding":277,"direction":120,"opportunity":278},"构建集成作物推荐、产量预测、肥料建议与多语言问答的AI农业决策Web平台。","随机森林、XGBoost、FastAPI、React PWA、OpenWeath","各模块准确率约88-95%，API、认证、翻译与聊天机器人测试均通过。","可引入真实田间时序数据与可解释性，验证模型跨区域跨季节泛化能力。","2026-10-07T23:30:47.248380Z",{"id":281,"title":282,"url":283,"summary":284,"summary_zh":9,"content":9,"source_name":285,"source_url":9,"published_at":286,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":287,"score_detail":288,"sources":290,"tags":292,"primary_topic":294,"secondary_topics":295,"classification":297,"search_phrases":318,"slug":321,"view_count":95,"doi":9,"paper":322,"created_at":330},4286,"Applications of Artificial Intelligence in Soil Characterization and Agriculture: A Systematic Review","https:\u002F\u002Fm2.mtmt.hu\u002Fapi\u002Fpublication\u002F37386408","Navarro Rubio等系统综述2018-2026年196项AI土壤特性与农业应用研究，遵循PRISMA 2020框架。综述发现研究明显向数据驱动方法转变，机器学习和深度学习模型主导近期研究；随机森林、支持向量机、梯度提升方法、人工神经网络、卷积神经网络、长短期记忆架构最常被报告；主要数据源包括原位传感、实验室测量、遥感影像和环境协变量，常通过多源数据融合框架集成；树集成模型在不同土壤特性上提供稳健性能。","Agronomy 2026 16(13):1241 DOI:10.3390\u002Fagronomy16131241","2026-10-02T00:00:00Z",78,{"impact":17,"substance":18,"depth":17,"authority":227,"freshness":125,"relevant":21,"comment":289},"PRISMA框架下196项研究的系统综述，方法学扎实、结论可靠，对农业AI土壤应用具较高参考价值。",[291],{"name":285,"url":283},[26,233,28,177,293],"土壤监测",{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},[296],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":45,"locked":14,"version":46,"method":47,"evidence":298},{"candidates":299,"agricultural_evidence":52},[300,305,310,314],{"code":39,"score":137,"supported":52,"title":301,"abstract":303,"tags":304},[302],"artificial intelligence",[66,28],[233,28],{"code":32,"score":306,"supported":52,"title":307,"abstract":308,"tags":309},11,[33],[55],[],{"code":174,"score":95,"supported":52,"title":311,"abstract":312,"tags":313},[],[177],[177],{"code":58,"score":21,"supported":14,"title":315,"abstract":316,"tags":317},[],[],[26],[319,320],"AI 土壤特性 系统综述","机器学习 土壤 遥感","AI土壤特性系统综述-4286",{"doi":9,"openalex_id":9,"authors":323,"venue":9,"cited_by_count":115,"oa_url":9,"card":324,"direction":120,"ingested_from":329},[],{"tldr":325,"method":326,"finding":327,"direction":120,"opportunity":328},"系统综述2018-2026年196项AI用于土壤特性与农业的研究。","PRISMA 2020框架，分析机器学习与深度学习模型及多源数据融合。","研究转向数据驱动，树集成模型在不同土壤特性上表现稳健。","可探索可解释AI与多源数据融合在土壤特性预测中的泛化能力。","agent","2026-10-06T00:05:04.976398Z",{"id":332,"title":333,"url":334,"summary":335,"summary_zh":336,"content":9,"source_name":337,"source_url":334,"published_at":338,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":339,"score_detail":340,"sources":344,"tags":346,"primary_topic":349,"secondary_topics":350,"classification":356,"search_phrases":380,"slug":383,"view_count":384,"doi":385,"paper":386,"created_at":404},3916,"Hyperspectral inversion of soil total nitrogen content using spectral and machine learning","https:\u002F\u002Fdoi.org\u002F10.3329\u002Fbjb.v55i3.93850","The Agro-pastoral Transitional Ecotone of Northern Shaanxi is an ecologically fragile region. Soil nitrogen is a critical factor influencing crop productivity and regional ecological balance. Rapid and accurate monitoring of farmland soil total nitrogen (TN) content is therefore essential for implementing precision agriculture and ensuring sustainable land use in this area. This study focused on the cultivated soils of Dingbian County, a representative area within this ecotone. Field-collected soil samples underwent TN content determination and laboratory-based hyper spectral reflectance measurement. The original spectral reflectance (R) was first transformed using mathematical methods, including first-derivative (FDR), second-derivative (SDR), and logarithm of the reciprocal's first-derivative (LRFD) analysis, to enhance spectral information and suppress background noise. Subsequently, the successive projections algorithm (SPA) and correlation analysis (CA) were integrated to identify feature wavelengths highly correlated with soil TN from the transformed spectral data. Partial Least Squares Regression (PLSR), Decision Tree (DT), and Random Forest (RF) were developed to estimate soil TN content. Model performance was evaluated and compared using the coefficient of determination (R²), root mean square error (RMSE), and ratio of performance to deviation (RPD). The results indicated that: (1) Spectral derivative transformations effectively highlighted subtle information, with LRFD showing the greatest sensitivity to TN. Models built using LRFD-derived features consistently outperformed those based on original reflectance. (2) Among the machine learning algorithms, the Random Forest (RF) model demonstrated superior predictive performance on the validation set (R² = 0.85, RMSE = 0.11 g\u002Fkg, RPD = 2.15), indicating the highest stability and generalization capability. The PLSR model performed moderately well, while the DT model was prone to over fitting and yielded the lowest accuracy. (3) This study confirms that the integration of hyper spectral technology and machine learning algorithms enables efficient and precise estimation of soil TN content in farmland within the Northern Shaanxi agro-pastoral ecotone, providing a reliable technical framework for soil fertility assessment and precision nutrient management in the region. Bangladesh J. Bot. 55(3): 605-614, 2026 (September)","陕北农牧交错带是生态脆弱区，土壤氮素是影响作物生产力和区域生态平衡的关键因子。快速、准确地监测农田土壤全氮（TN）含量，对于该区实施精准农业和保障土地可持续利用至关重要。本研究以该交错带典型区域定边县的耕作土壤为研究对象，对野外采集的土壤样品进行TN含量测定和实验室高光谱反射率测量。首先采用数学方法对原始光谱反射率（R）进行变换，包括一阶微分（FDR）、二阶微分（SDR）和倒数对数一阶微分（LRFD）分析，以增强光谱信息并抑制背景噪声。随后，将连续投影算法（SPA）与相关分析（CA）相结合，从变换后的光谱数据中筛选出与土壤TN高度相关的特征波长。构建偏最小二乘回归（PLSR）、决策树（DT）和随机森林（RF）模型估算土壤TN含量。采用决定系数（R²）、均方根误差（RMSE）和相对分析误差（RPD）对模型性能进行评价与比较。结果表明：（1）光谱微分变换有效凸显了细微信息，其中LRFD对TN的敏感性最强，基于LRFD衍生特征构建的模型始终优于基于原始反射率的模型。（2）在机器学习算法中，随机森林（RF）模型在验证集上表现出优越的预测性能（R² = 0.85，RMSE = 0.11 g\u002Fkg，RPD = 2.15），具有最高的稳定性和泛化能力。PLSR模型表现中等，而DT模型易于过拟合，精度最低。（3）本研究证实，高光谱技术与机器学习算法相结合能够高效、精确地估算陕北农牧交错带农田土壤TN含量，为该区域土壤肥力评估和精准养分管理提供了可靠的技术框架。Bangladesh J. Bot. 55(3): 605-614, 2026 (September)","Bangladesh Journal of Botany","2026-09-30T00:00:00Z",70,{"impact":341,"substance":342,"depth":139,"authority":19,"freshness":20,"relevant":21,"comment":343},12,20,"高光谱结合机器学习反演土壤全氮，方法对比扎实、结论可靠，对精准农业养分管理有参考价值，但属区域性田间研究，影响范围有限。",[345],{"name":337,"url":334},[26,347,28,177,348],"精准施肥","土壤氮素",{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},[351,352],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"code":174,"slug":353,"name":354,"path":355,"parent_code":43},"remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing",{"status":45,"locked":14,"version":46,"method":47,"evidence":357},{"candidates":358,"agricultural_evidence":52},[359,363,367,372,376],{"code":39,"score":51,"supported":52,"title":360,"abstract":361,"tags":362},[67],[28,67],[28],{"code":32,"score":227,"supported":52,"title":364,"abstract":365,"tags":366},[33],[55,33],[],{"code":174,"score":228,"supported":52,"title":368,"abstract":370,"tags":371},[369],"hyperspectral",[],[177],{"code":58,"score":72,"supported":52,"title":373,"abstract":374,"tags":375},[],[29,60],[26],{"code":71,"score":95,"supported":52,"title":377,"abstract":378,"tags":379},[],[76],[347],[381,382],"定边县 土壤全氮 高光谱","陕北农牧交错带 土壤氮素 反演","定边县土壤全氮高光谱-3916",2,"10.3329\u002Fbjb.v55i3.93850",{"doi":385,"openalex_id":387,"authors":388,"venue":337,"cited_by_count":115,"oa_url":334,"card":399,"direction":120,"ingested_from":122},"W7214885868",[389,392,394,396],{"name":390,"orcid":391},"Zhenyu Dong","https:\u002F\u002Forcid.org\u002F0000-0003-0028-3293",{"name":393,"orcid":9},"Ni Wang",{"name":395,"orcid":9},"Jiancang Xie",{"name":397,"orcid":398},"Jungang Luo","https:\u002F\u002Forcid.org\u002F0009-0009-0285-4839",{"tldr":400,"method":401,"finding":402,"direction":213,"opportunity":403},"用高光谱与机器学习反演陕北农牧交错带农田土壤全氮含量。","野外采样测光谱，经FDR\u002FSDR\u002FLRFD变换，SPA+CA选特征波长，建PLS","LRFD变换最敏感，RF模型验证集R²=0.85、RPD=2.15，精度与泛化最优。","可探索多源遥感与深度学习融合，提升生态脆弱区土壤氮素时空反演与制图能力。","2026-10-01T23:30:53.960449Z",{"id":406,"title":407,"url":408,"summary":409,"summary_zh":410,"content":9,"source_name":411,"source_url":408,"published_at":412,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":413,"score_detail":414,"sources":416,"tags":418,"primary_topic":420,"secondary_topics":421,"classification":423,"search_phrases":448,"slug":451,"view_count":384,"doi":452,"paper":453,"created_at":477},3805,"Machine Learning Prediction of Soil Nutrients and pH in Tibetan Alpine Croplands: A Comparison of Nine Algorithms","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fsu18199875","Soils underpin terrestrial ecosystem functioning and exert strong control over farmland productivity, nutrient turnover, environmental quality, and agricultural sustainability. Reliable prediction of soil nutrients and pH is therefore important for precision fertilization, soil-quality assessment, and the sustainability of farmland management, particularly in fragile alpine agroecosystems. However, the comparative performance of candidate algorithms for estimating soil properties in alpine croplands of the Qinghai-Tibetan Plateau remains poorly characterized. Here, we assessed nine approaches, namely random forest (RF), generalized boosting regression (GBR), multiple linear regression (MLR), artificial neural network (ANN), generalized linear regression (GLR), conditional inference tree (CIT), extreme gradient boosting (eXGB), support vector machine (SVM), and recursive partitioning regression tree (RRT), for predicting soil carbon, nitrogen, phosphorus, potassium, and pH from annual mean temperature, annual precipitation, annual solar radiation, and maximum normalized difference vegetation index (NDVImax). Field observations were obtained from 103 sampling plots, with four individual soil samples collected per plot and depth, resulting in 412 observations for each depth layer. Clear contrasts in model performance emerged under alpine agricultural conditions. Across all soil variables and depths, eXGB and RF showed the best overall performance, with the lowest mean relative bias values of 1.35% and 1.55%, respectively, and the lowest mean RMSE values of 0.43 and 0.47. These results indicate that nonlinear ensemble algorithms better captured the heterogeneous relationships between environmental covariates and soil properties than linear models or single-tree approaches. Overall, RF and eXGB provide a practical methodological basis for precision fertilization, soil-quality assessment, and long-term sustainability in fragile alpine agroecosystems.","土壤是陆地生态系统功能的基础，并对农田生产力、养分周转、环境质量和农业可持续性具有强烈控制作用。因此，可靠预测土壤养分和pH对于精准施肥、土壤质量评估以及农田管理的可持续性十分重要，尤其是在脆弱的高寒农业生态系统中。然而，在青藏高原高寒农田中，用于估算土壤属性的候选算法的比较性能仍缺乏系统表征。本研究评估了九种方法，即随机森林（RF）、广义提升回归（GBR）、多元线性回归（MLR）、人工神经网络（ANN）、广义线性回归（GLR）、条件推断树（CIT）、极端梯度提升（eXGB）、支持向量机（SVM）和递归分割回归树（RRT），利用年平均气温、年降水量、年太阳辐射和最大归一化差异植被指数（NDVImax）预测土壤碳、氮、磷、钾和pH。野外观测来自103个采样样地，每个样地和每个深度采集4个独立土壤样品，每个深度层共获得412个观测值。在高寒农业条件下，模型性能表现出明显差异。在所有土壤变量和深度上，eXGB和RF总体表现最佳，平均相对偏差值最低，分别为1.35%和1.55%，平均RMSE值最低，分别为0.43和0.47。这些结果表明，与线性模型或单树方法相比，非线性集成算法更好地捕捉了环境协变量与土壤属性之间的异质性关系。总体而言，RF和eXGB为脆弱高寒农业生态系统中的精准施肥、土壤质量评估和长期可持续性提供了实用的方法学基础。","Sustainability","2026-09-27T00:00:00Z",73,{"impact":341,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":415},"基于103个样点412份观测对比九种算法预测高原农田土壤养分与pH，方法扎实、结论明确，对精准施肥有直接参考价值，但属区域性研究，产业影响有限。",[417],{"name":411,"url":408},[26,233,347,30,419],"青藏高原",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[422],{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},{"status":45,"locked":14,"version":46,"method":47,"evidence":424},{"candidates":425,"agricultural_evidence":52},[426,430,434,440,444],{"code":32,"score":51,"supported":52,"title":427,"abstract":428,"tags":429},[33],[55,33],[30],{"code":39,"score":228,"supported":52,"title":431,"abstract":432,"tags":433},[67],[],[233],{"code":71,"score":72,"supported":52,"title":435,"abstract":436,"tags":439},[],[437,438],"施肥","fertilization",[347],{"code":79,"score":80,"supported":52,"title":441,"abstract":442,"tags":443},[],[83],[],{"code":58,"score":21,"supported":14,"title":445,"abstract":446,"tags":447},[],[],[26],[449,450],"青藏高原 农田 土壤养分 机器学习","土壤pH 预测 随机森林 XGBoost","青藏高原农田土壤养分机器学习-3805","10.3390\u002Fsu18199875",{"doi":452,"openalex_id":454,"authors":455,"venue":411,"cited_by_count":115,"oa_url":408,"card":472,"direction":120,"ingested_from":122},"W7214680092",[456,458,460,463,465,468,470],{"name":457,"orcid":9},"Chenjun Zhao",{"name":459,"orcid":9},"Shaowei Li",{"name":461,"orcid":462},"Yuan Tian","https:\u002F\u002Forcid.org\u002F0000-0002-8758-1826",{"name":464,"orcid":9},"ChengQun Yu",{"name":466,"orcid":467},"Gang Fu","https:\u002F\u002Forcid.org\u002F0000-0002-6055-8730",{"name":469,"orcid":9},"Zhiming Zhong",{"name":471,"orcid":9},"Wei Sun",{"tldr":473,"method":474,"finding":475,"direction":120,"opportunity":476},"比较九种机器学习算法预测青藏高原高寒农田土壤养分与pH，发现集成算法最优。","基于103个样点412个观测，用气候与NDVI变量比较RF、eXGB等九种算法。","eXGB和RF表现最佳，平均相对偏差最低（1.35%和1.55%），优于线性与单树模型。","可引入多时相遥感与深度学习方法，提升高寒区土壤属性制图精度与泛化能力。","2026-09-29T23:30:57.794777Z",{"id":479,"title":480,"url":481,"summary":482,"summary_zh":483,"content":9,"source_name":484,"source_url":481,"published_at":412,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":413,"score_detail":485,"sources":487,"tags":489,"primary_topic":492,"secondary_topics":495,"classification":498,"search_phrases":530,"slug":533,"view_count":384,"doi":534,"paper":535,"created_at":548},3804,"Evaluating Reference Evapotranspiration and Soil Moisture as Predictors for Machine Learning-Based Actual Evapotranspiration Estimation: Model Performance and Explainability","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fatmos17100941","Accurate estimation of actual evapotranspiration (ETa) is essential for sustainable water resources management and agricultural planning. This study investigated how the availability of reference evapotranspiration (ET0) and soil moisture (SM) affects ETa prediction performance and predictor importance using Random Forest (RF), Bagging Trees (BT), Least Squares Boosting (LSBoost), Generalized Additive Models (GAM), and Multiple Linear Regression (MLR). Global solar radiation (Rs), land surface temperature (LST), normalized difference vegetation index (NDVI), and SM were evaluated under two primary scenarios, with and without ET0. Models were developed using 80% of the data, with five-fold cross-validation conducted within the training subset, and final performance was evaluated on an independent 20% test subset. Model interpretability was assessed using SHAP, Permutation Feature Importance (PFI), Partial Dependence Plot (PDP), and temporal signed-SHAP analyses, while an additional SM-ablation experiment quantified the incremental predictive contribution of soil moisture. In Scenario I, ET0 was the dominant predictor, and RF and BT achieved the highest independent-test performance (R2 = 0.875; RMSE = 0.398–0.399 mm day−1). Excluding ET0 reduced predictive performance, with RF, LSBoost, and BT achieving R2 values of approximately 0.80 in Scenario II, while the predictor-importance structure shifted primarily toward Rs, followed by SM, LST, and NDVI. Removal of SM further reduced performance in both scenarios, with substantially greater deterioration when ET0 was unavailable: R2 decreased by 0.025–0.044 in Scenario I and by 0.051–0.099 in Scenario II. Under the SM-excluded Scenario I configuration, which more closely aligned the predictor information with previous ANN- and SAFER-based studies, R2 values ranged from 0.820 to 0.832, indicating that differences in ETa prediction performance cannot be attributed solely to model structure but also depend strongly on predictor information content. Overall, ET0 provides substantial atmospheric-demand information for ETa prediction, whereas Rs assumes the dominant predictive role, and SM provides particularly important complementary information when ET0 is unavailable. The combined performance, ablation, and explainability analyses demonstrate the importance of considering predictor composition together with model structure when developing and comparing ETa estimation approaches.","准确估算实际蒸散发（ETa）对于可持续水资源管理和农业规划至关重要。本研究探讨了参考蒸散发（ET0）和土壤水分（SM）的可用性如何影响ETa预测性能及预测因子重要性，采用了随机森林（RF）、Bagging Trees（BT）、最小二乘提升（LSBoost）、广义加性模型（GAM）和多元线性回归（MLR）方法。在全球太阳辐射（Rs）、地表温度（LST）、归一化差异植被指数（NDVI）和SM的条件下，评估了包含和不包含ET0两种主要情景。模型使用80%的数据进行开发，在训练子集内进行五折交叉验证，最终性能在独立的20%测试子集上评估。模型可解释性通过SHAP、排列特征重要性（PFI）、偏依赖图（PDP）和时间符号SHAP分析进行评估，同时通过额外的SM消融实验量化了土壤水分的增量预测贡献。在情景I中，ET0是主导预测因子，RF和BT取得了最高的独立测试性能（R2 = 0.875；RMSE = 0.398–0.399 mm day−1）。排除ET0降低了预测性能，在情景II中RF、LSBoost和BT的R2值约为0.80，同时预测因子重要性结构主要转向Rs，其次是SM、LST和NDVI。去除SM进一步降低了两种情景下的性能，且在ET0不可用时性能下降幅度更大：情景I中R2下降了0.025–0.044，情景II中下降了0.051–0.099。在排除SM的情景I配置下（该配置使预测因子信息更接近此前的ANN和SAFER研究），R2值范围为0.820至0.832，表明ETa预测性能的差异不能仅归因于模型结构，还强烈依赖于预测因子信息含量。总体而言，ET0为ETa预测提供了大量大气需求信息，而Rs承担了主导预测作用，SM在ET0不可用时提供了特别重要的补充信息。综合性能、消融和可解释性分析表明，在开发和比较ETa估算方法时，需将预测因子组成与模型结构一并考虑。","Atmosphere",{"impact":341,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":486},"方法对比与可解释性分析扎实，对农业水资源管理和灌溉决策有参考价值，但属专业论文，公共影响有限。",[488],{"name":484,"url":481},[26,490,28,145,491],"农业遥感","蒸散发估算",{"code":182,"slug":186,"name":493,"path":494,"parent_code":36},"水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation",[496,497],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},{"status":45,"locked":14,"version":46,"method":47,"evidence":499},{"candidates":500,"agricultural_evidence":52},[501,508,512,516,522,526],{"code":182,"score":139,"supported":52,"title":502,"abstract":504,"tags":507},[503],"evapotranspiration",[505,506,503],"水资源","蒸散发",[],{"code":32,"score":51,"supported":52,"title":509,"abstract":510,"tags":511},[33],[55,33],[145],{"code":39,"score":228,"supported":52,"title":513,"abstract":514,"tags":515},[67],[],[28],{"code":517,"score":80,"supported":52,"title":518,"abstract":519,"tags":521},"J1",[],[520],"农业规划",[],{"code":58,"score":21,"supported":14,"title":523,"abstract":524,"tags":525},[],[],[26],{"code":174,"score":21,"supported":14,"title":527,"abstract":528,"tags":529},[],[],[490],[531,532],"参考蒸散发 土壤水分 机器学习","ET0 实际蒸散发 随机森林","参考蒸散发土壤水分机器学习-3804","10.3390\u002Fatmos17100941",{"doi":534,"openalex_id":536,"authors":537,"venue":484,"cited_by_count":115,"oa_url":481,"card":543,"direction":120,"ingested_from":122},"W7214588993",[538,541],{"name":539,"orcid":540},"Halil Karahan","https:\u002F\u002Forcid.org\u002F0000-0001-5346-5686",{"name":542,"orcid":9},"Devrim Alkaya",{"tldr":544,"method":545,"finding":546,"direction":120,"opportunity":547},"评估ET0与土壤湿度对机器学习估算实际蒸散的影响及可解释性。","用RF、BT、LSBoost、GAM、MLR结合SHAP、PFI、PDP及SM消","ET0主导预测，缺失时Rs主导且SM补充价值显著增大。","可探索缺ET0条件下融合遥感与土壤湿度的可解释ETa估算框架。","2026-09-29T23:30:57.729880Z"]