深度学习赋能深层地热资源勘探:进展与趋势

崔哲思 ,  黄学莲 ,  蒋恕 ,  王洋 ,  王昭君 ,  彭昊 ,  王帅 ,  陈麒玉 ,  刘刚

地球科学 ›› 2026, Vol. 51 ›› Issue (03) : 1144 -1164.

PDF (3327KB)
地球科学 ›› 2026, Vol. 51 ›› Issue (03) : 1144 -1164. DOI: 10.3799/dqkx.2026.017

深度学习赋能深层地热资源勘探:进展与趋势

作者信息 +

Deep Learning Empowers Deep Geothermal Resources Exploration: Progress and Trend

Author information +
文章历史 +
PDF (3406K)

摘要

全球深层地热资源勘探开发正处于从“实验性探索”向“规模化应用”转型的关键阶段.以深度学习为代表的人工智能已经在大数据分析、模式识别和非线性问题求解方面展现出变革性潜力,为破解当前制约我国深层地热资源高效精准勘探的核心难题提供了新途径.推动深度学习与传统地热勘探流程的深度融合,对我国在全球深层地热资源开发利用领域提升竞争力具有重要意义.本文聚焦深度学习数据处理、建模及预测技术与深层地热资源勘探(涵盖地热地质勘查、地球物理勘探、地球化学勘探等环节)的深度结合,系统梳理并总结了地热资源勘探技术、深度学习技术方法、深度学习赋能深层地热资源勘探的关键技术方法、核心进展与研究成果,展现了深度学习赋能地热资源勘探方法相较于传统方法带来的效率、准确率与精度提升.本文最后结合前沿技术阐述了深度学习赋能深层地热资源勘探领域面临的核心挑战,未来深层地热资源智能勘探亟须聚焦多模态数据融合、可解释与可信人工智能、地热垂直领域智能计算基座与大模型建设等方面,最终实现从“经验驱动”到“知识驱动”再到“智能驱动”的跨越,为地热能源资源行业数字化智能化发展提供核心技术支撑.

Abstract

The global exploration and development of deep geothermal resources is at a critical stage, transitioning from experiment to application. Artificial intelligence, particularly deep learning, has demonstrated transformative potential in big data analysis, pattern recognition, and nonlinear problem solution, offering new pathways to address challenges hindering efficient and precise exploration of deep geothermal resources. It is significantly important to promote the integration of deep learning with traditional geothermal exploration processes to enhance China’s competitiveness in the development and utilization of deep geothermal resources. This paper focuses on the integration of deep learning data processing, modeling, and prediction with deep geothermal resource exploration (including geothermal geological surveys, geophysical exploration, and geochemical exploration, etc.). It systematically reviews and summarizes key technological methods in geothermal resource exploration, deep learning techniques, and the critical advancements and research outcomes that empower deep geothermal exploration. This study demonstrates the improvement of efficiency, accuracy, and precision brought by deep-learning-based geothermal resources exploration methods compared to traditional methods. Finally, the paper discusses the core challenges with cutting-edge technologies faced by deep geothermal exploration. In future, intelligent deep geothermal resources exploration urgently needs to focus on multiple modal data fusion, interpretable and trustworthy artificial intelligence, and construction of intelligent computing foundations and large models, which ultimately, will enable a leap from “experience-driven” to “knowledge-driven” and then to “intelligent-driven”, providing core technological support for the digital and intelligent development of the geothermal energy resources industry.

Graphical abstract

关键词

深层地热资源 / 地热资源勘探 / 人工智能 / 深度学习.

Key words

deep geothermal resources / geothermal resources exploration / artificial intelligence / deep learning

引用本文

引用格式 ▾
崔哲思,黄学莲,蒋恕,王洋,王昭君,彭昊,王帅,陈麒玉,刘刚. 深度学习赋能深层地热资源勘探:进展与趋势[J]. 地球科学, 2026, 51(03): 1144-1164 DOI:10.3799/dqkx.2026.017

登录浏览全文

4963

注册一个新账户 忘记密码

0 引言

在全球能源格局深刻变革、国际政治经济局势复杂多变的背景下,我国面临能源结构转型及绿色发展等诸多挑战.地热能作为重要的非碳基可再生资源,因其具备本土能源、稳定可靠、绿色低碳等特质,展现出显著优势 (王贵玲等, 2020).21世纪以来,我国地热直接利用规模稳居世界首位,中深层地热供暖利用快速增长,已经成为全球地热直接利用规模最大的国家 (孙焕泉等, 2024).然而,与浅层地热的广泛应用、中深层地热利用逐步稳定增长的现状形成鲜明对比的是,我国深层地热资源的勘探开发仍处于起步和探索阶段 (庞忠和等, 2020).深层地热资源通常埋藏于3~10 km甚至更深的地壳中,包括了干热岩以及在特定深度和温压条件下存在的超临界热流体,具有隐蔽性强、温度高、储量大、但可获取性差等典型特征 (王贵玲和陆川, 2022),其成藏地质过程极为复杂,往往经历了多期次的构造运动、热液活动及沉积环境变化,最终形成强非均质、显著各向异性的资源空间分布.且深层地热储层埋深大、勘探成本高,地震、电磁及地球化学观测数据质量较低,数据噪声强,传统地质手段与数值反演方法难以实现对深部热储结构、温度场及流体分布的高精度刻画,导致深层高品位地热资源难以被准确预测,制约了深层地热资源的高效开发 (李丛等, 2021).

人工智能(Artificial Intelligence, AI)的快速发展为深层地热的勘探与开发提供了前所未有的机遇 (匡立春等, 2021; 李根生等, 2024a; 王天宇等, 2024).人工智能是一种通过计算机模拟人类认知过程、实现数据感知、模式识别与智能决策的综合性技术体系,其核心思想是从大量数据中自动提取潜在特征与规律,实现非线性、高维、复杂系统的高效建模与推理.作为AI领域的核心分支,深度学习通过利用包含大量神经元的多层神经网络模拟人脑学习方式和认知过程,已经在跨尺度数据融合、复杂非线性与高维问题求解、实时学习与动态优化等方面展现了卓越性能 (Lecun et al., 2015).

目前,以深度学习为代表的人工智能技术已经在油气、矿产等各类地质资源预测和建模领域发挥了重要作用(成秋明, 2021, 2025;王国法等, 2021; 周永章等, 2021).这一智能化的发展趋势也已经在国际地热勘探领域得到应用验证.2025年9月,美国地热勘探企业Zanskar利用人工智能技术成功预测并钻获了内华达州北部Pumpernickel地区第二处高温地热资源,该地热井在井深762 m处温度已达到137 ℃,验证了AI在数据稀疏地热田识别高潜力靶区的能力 (Carlo, 2025).当前,我国正在加速推进“人工智能+”能源战略 (http://www.scio.gov.cn/zdgz/jj/202509/t20250901_928364.html),2025年9月国家发展改革委、国家能源局印发的《关于推进“人工智能+”能源高质量发展的实施意见》(https://www.gov.cn/zhengce/zhengceku/202509/content_7040253.htm)中明确要求,到2027年初步构建能源与人工智能融合创新体系,在能源资源智能勘探等方向取得突破;到2030年相关技术达到世界领先水平.因此,将人工智能尤其是深度学习与深层地热资源勘探融合,不仅是当前地热新能源领域发展的前沿方向,也是突破深层地热资源隐蔽性强、数据稀疏、系统耦合复杂等关键瓶颈的有效途径.通过推动深度学习与地热资源勘探的深度融合,有望利用深度学习的跨尺度数据感知能力、多模态数据融合能力与因果关联推理能力,高效利用地质、地球物理、地球化学等多源多模态地热数据,实现对深层地热资源靶区、深层热储结构、热储参数以及地热场动态演化等的精准预测和识别,进而突破当前制约我国深层地热资源高效精准勘探的核心难题,推动我国在全球深层地热开发利用领域走在前列.

为此,笔者将从深层地热资源勘探技术手段与研究现状、深度学习技术方法与研究现状、深度学习赋能深层地热资源勘探研究实例等不同角度总结和概括深度学习赋能深层地热资源勘探技术的最新研究进展,并对前沿交叉技术挑战与未来发展趋势给出了笔者的理解和认识.

1 地热资源勘探技术方法与研究现状

地热资源勘探技术方法包括了地质与地表勘探、地球物理勘探、地球化学勘探、地质工程钻探等多个环节.如图1所示,通过地质与地表勘探技术方法识别地热活动痕迹与线索,在此基础上通过地球物理与地球化学勘探手段进行数据采集和分析,进而预测潜在的地热资源勘探靶区,最终通过地质工程钻探进行验证.笔者将对传统地热资源勘探技术方法的流程和研究现状进行分析和总结.

1.1 地质与地表勘探

地表勘探通常依据遥感数据对研究区开展资源初步筛查,由于遥感技术能够对地表特征进行大规模测绘,可以此作为地热资源识别的重要手段,且具有速度快、时效性强、范围广、成本低、受地表条件限制小等优势 (周桃勇等, 2020; 辛磊等, 2021).早在20世纪70年代,遥感技术便被应用于辽南、天津、福建等地区开展地热调查,并在辽南地区取得了明显的应用效果(霍超等, 2023).短波红外(SWIR)和可见光(VNIR)波段可揭示与地热流体和热通量等属性相关的次生矿物蚀变区域(Afshar et al., 2023),但受分辨率、成像频度等的限制,依靠单一热红外探测数据来揭示地下构造和流体通道往往效果不佳 (Moradi et al., 2015).热红外遥感通过反演地表温度,能够识别与地下热活动相关的地表温度异常区域,这种方法对区域性地热远景区的初步圈定具有应用价值.野外地质调查能够对地表地质现象进行系统观察和测量,可直接识别热流体运移通道、热储层与盖层等核心要素.典型地热显示包括温泉、蒸汽口、泥浆池、硅华沉积、黄铁矿化带及热液蚀变带 (陈墨香, 1992; 丛婷婷等, 2024).高孔隙度、高渗透性的火山碎屑岩、玄武岩裂隙体、砂岩及碳酸盐岩溶蚀体系通常构成优质储层 (王贵玲等, 2023).盖层的判识则强调岩性致密度与构造完整性.泥岩、页岩、凝灰岩、致密火山岩及部分变质岩因渗透率低、延展性强且分布连续,是优质盖层(周雁等, 2011).通过结合遥感识别的地表异常与野外调查获取的构造形迹、岩性分布、裂隙网络及热液蚀变信息,可有效判断热储、盖层及流体通道特征,建立地热系统的初步地质框架,这些基础数据为后续物探、化探与钻探部署提供了关键约束.但地表勘探手段在探测深层地热资源时,地表热异常信号可能非常微弱,其直接指示深部地热潜力的能力有限.

1.2 地球物理勘探

地球物理技术主要利用地震波和电磁波的反射、局部重力和磁场的变化以及热梯度等方法表征地热系统的裂缝网络、断层、岩性变化、热通量、流体的存在和渗透率边界等结构信息和特征,常用方法主要有地震勘探、电磁/电法勘探以及重/磁勘探等(Qiao and Zhang, 2025).

地震勘探方法旨在通过地震反射波来解析地下结构、断裂、界面与孔隙性地层等,进而识别地下流体/裂缝系统,在构造复杂区和高孔隙体‒断层区可结合速度/阻抗反演综合解释,明确热储地区构造空间展布,具有勘探深度大、地层分辨率高等优点 (陈昌昕等, 2020; 常宽等, 2025).Piana Agostinetti et al. (2017) 利用地震勘探技术发现了位于Larderello地热田中心 4 km深处的超临界流体地层.Fu et al. (2022) 通过对地震数据进行去噪和弱深有效信号处理,能够有效实现干热岩储层的高质量成像,并识别热源、储层、通道和盖层.地震微动方法利用地球本身的微弱震动来刻画地下介质构造,已被广泛应用于浅部地热资源调查 (董耀等, 2020; 田宝卿等, 2020).但此类方法探测精度与反射地震方法相比较低,难以适用于深层地热资源勘查.磁法勘查主要用于识别断裂位置及圈定岩体范围,在干热岩勘探中被主要用于确定隐伏火成岩体的分布、厚度及其与断裂带的关系(王丹凤和程剑峰, 2023).张森琦等(2020)基于高精度航磁测量数据的反演计算结果,圈定了共和盆地15处隐伏干热岩地热资源勘查目标靶区.电磁法/电法利用热源、储层、盖层等不同部分的温度差和电阻率差异对地下热储进行勘探(何继善, 2020; 覃田赐等,2020).Zhu et al.(2022)通过可控源音频大地电磁法(CSAMT)与瞬变电磁法(TEM)的组合,成功识别了控热断裂与热储边界,CSAMT的高分辨率特性可精细刻画浅部热储结构,而TEM对深部高导体的敏感性则补充了深部热源信息.吕清等(2025)基于CSAMT三维反演技术,成功实现了浙东火山岩区的深部地热构造勘查.深部的超临界地热流体往往具有极高的导电性,这使得电磁法在超临界流体勘探领域具有较好的应用前景 (Yamaya, 2025).Ishizu et al. (2022) 通过采集的大地电磁数据揭示了位于Yuzawa 地热田2.5~6.0 km深处的潜在超临界地热储层.重力勘探法是利用由岩/矿石密度差异引起的重力变化而进行地质勘查的一种方法,在研究地球深部构造、区域地质构造、隐伏构造以及地层分布规律方面具有重要作用,可用于识别构造尺度和岩性差异,具有成本低、不受磁电干扰、勘探深度大等优点(Ragueh et al., 2024).近年来,多方法综合地球物理勘探已成为地热资源勘查的重要手段,通过集成不同地球物理数据的互补性,显著提升了热储定位与成因机制解析的精度.Afshar et al. (2023) 通过联合重力、磁法和大地电磁等多种地球物理数据,预测了伊朗Sabalan地热田的热源、流体通道与储层的空间配置关系,验证了多方法协同在复杂地质背景下的适用性.然而,多方法地球物理数据融合仍缺乏标准化流程,且对深部热储存在的电阻率、密度、磁化率等多物性参数耦合响应仍缺乏有效的解耦方法.

1.3 地球化学勘探

岩石地球化学、水文地球化学和气体地球化学分析是地球化学地热资源勘探中较为常用的方法.岩石地球化学旨在通过分析泉华、水热蚀变矿物以及储层岩性等地化特征,并根据其矿物特征、微量元素特征等推测热气体、温度场、热流体通道、储层孔隙发育程度,进而判断深部化学环境、地热热源、地热流体化学迁移、富集及热循环路径 (赵钵渊等, 2024; 雷云开等, 2025).水文地球化学分析则通过对地热水的水化学组分及同位素特征,揭示地热流体的补给来源、形成时间、运移过程、演化规律等,还可以用水‒岩平衡来研究热储层温度及运移过程中的混合作用 (刘颖超等, 2015; 高宗军等, 2019),其内容主要包括温度、pH、电导率、碱度、微量元素及阴阳离子、同位素等指标的测试分析 (Bénard et al,2024; 卫兴等, 2024; 辛浩等, 2024).现有水文地球化学分析在地热勘查过程中的储层温度预测、水源识别、地热井预测结垢和腐蚀方面的应用效果显著 (郭清海,2020).气体地球化学勘查则通过分析温泉气、蒸汽孔气、土壤气(CO2、Rn等)、钻井井口气体以及溶解气等与热源相关的地热气体组分(He、Ne、Ar、Kr、Xe等稀有气体)及其同位素信息来探测深部热源,进而分析地热系统中不同热源的占比以及控热构造与深部的连通性 (孙占学等, 2004; 天娇等, 2022; Huang et al., 2024Wang et al., 2025a).在勘查实践中,岩石、水文和气体地球化学方法常常被综合应用以获得更可靠的结果.

1.4 地质工程钻探

地质工程钻探是地热勘探的核心与验证环节,该过程旨在通过钻孔直接揭示地下地层、断裂、岩性、温度及热储特征,为地热田的评价与开发提供最直接的数据支撑.钻探所得岩心样本、测井数据等可为孔渗分析、地温分析、储层评价和热能抽提系统设计提供数据基础,也可用于建立地热储层模型、评估热储产能和井筒生产特征,从而构成地热田评价与开发设计的基础 (雷玉德等, 2023; 岳雯秀等, 2024).在钻井过程中,会进行随钻录井、电测井以及井下压力、温度测量等地球物理测井工作,以记录地层物性参数以及井筒温度、压力等状况.钻井完钻后还要进行注水或抽水,以测定井产能指数和井系数等性能参数,并采集井筒热流体样品进行化学分析,上述测井和采样结果为热储类型判定和产能评估提供了关键依据 (Kristensen et al., 2016Salimzadeh et al., 2019; Rangel⁃Arista et al., 2025).现阶段我国钻探技术取得系列重大突破,中国首口超5 000 m深层地热科学探井——福深热1井正式完工标志着我国已具备深部地热资源钻探技术 (贠晓瑞等, 2026).然而地热钻井成本高昂,一般占地热开发总投资的30%~70%,因此提高钻探效率和精度对控制勘探成本、确保地热开发成功具有关键作用.

2 用于地热资源勘探的深度学习技术方法与研究现状

以深度学习为代表的人工智能技术已经成为多学科、多领域大数据挖掘、建模/模拟、分析与预测的核心和关键技术(Van Noorden and Perkel, 2023).深度学习方法具有强大的复杂空间模式抽象能力、深层特征提取和生成能力以及数据通路高效复用能力 (张国印等, 2024),能够为地热勘探评价开发全流程提供关键技术支撑.深度学习技术方法根据其学习方式,可分为监督学习、无监督学习、半监督学习和强化学习,不同学习类型的深度学习技术在地热资源勘探领域的应用方向不同(图2).例如,监督学习方法可有效实现热储品位预测,通过建立地温梯度、热储温度等关键参数与已知区域热储品位等级之间的映射关系,构建深度学习模型以提取控热要素与品位分级之间的非线性关联规律.无监督学习方法适用于热储物性参数隐含特征提取,其通过深层神经网络将已知探区物性参数及地质环境特征映射至高维隐蔽空间,随后可基于概率采样和近邻匹配算法重构物性参数的空间分布.半监督学习方法能够协同利用少量标记数据和大量未标记数据,在低勘探程度区域实现资源潜力评估.强化学习方法通过智能体与环境的交互反馈机制优化决策过程,被应用于深层地热钻探工程以提高优质深层热储钻遇概率.笔者将从深度学习技术的不同学习方式出发,分析并总结各类深度学习方法的基本理念和方法,并指出与地热资源勘探结合的方式.

2.1 监督学习

基于监督学习的深度学习方法旨在利用带标签的数据进行训练,并使用深度神经网络学习输入特征与输出标签之间的映射关系,最终实现对未知数据的预测或分类(周飞燕等, 2017).在监督学习范式下,输入特征就代表了描述样本的多维可观测属性,而输出标签则对应期望的地质目标或实际观测值.以地热资源分布预测为例(如图3所示),输出标签可以是地热资源的空间分布状态,输入特征则可涵盖断裂构造分布、储层展布特征、地层厚度、地球物理场信息、温度场数据等多种类型、多源、多维的特征参数(蒋恕等, 2020).通过输入特征和输出标签构建的训练样本监督深度神经网络训练过程,使深度神经网络提取输入特征到输出标签的隐含非线性相关关联,从而建立可用于推断未探区地热资源潜力的高精度预测模型.按照输出预测结果的数值类型,可将监督学习划分为分类方法和回归方法.分类方法的输出为离散型地质变量,例如岩性识别、岩相划分、储层类型判别、资源类别判定等;回归方法则用于预测连续型地质参数,如地温梯度、热导率、孔隙度、渗透率、热储层温度等地热关键属性(Assouline et al., 2019Zheng et al., 2025).

2.2 无监督学习

无监督学习不依赖于预先定义的标签,其核心目标是从数据本身发掘其内在的结构、分布或代表性特征(殷瑞刚等, 2016).在地热资源勘探领域,当地热勘探区域缺乏充足的已钻井数据或已知的地热异常标签时,基于深度学习聚类算法能够基于区域断裂、地质资料、布格重力异常、航磁异常等地球物理场、地球化学元素浓度及遥感光谱特征等多元数据集,将研究区划分为若干具有相似地球物理‒化学属性的子区域,以揭示潜在的热储构造边界、圈定热液蚀变带或识别与地热活动相关的隐伏岩体(Wang et al., 2023cPalo et al., 2024).基于无监督学习的聚类算法能够对高维、冗余的地球物理叠后数据或遥感影像进行特征压缩、提取与解译(Li et al., 2023),从而滤除噪声干扰,凸显与地热系统相关的核心地质要素 (Chopra et al., 2021),为后续的靶区优选提供定量化依据(图4所示).

2.3 半监督学习

半监督学习结合少量标注数据和大量未标注数据共同训练神经网络模型,这一特性使其非常契合地热勘探中“已知点少、未知区域广”的普遍现状,常用方法包括伪标签法和一致性正则化.伪标签法特别适合地热领域中标签稀缺,但数据样本丰富的实际情况(Guo et al., 2024),而一致性正则化法则通过对输入数据施加轻微扰动,并约束模型在此扰动下保持预测结果的一致性,从而充分利用大量未标注数据.在地热资源勘探场景下,伪标签法可利用已钻井的测井数据(有标签)与区域地球物理勘探数据(无标签)联合训练模型,提升对储层属性或热异常的识别能力,从而实现在降低人工标注成本的同时,有效提高模型的泛化性能与稳健性(Asghar et al., 2020).一致性正则化方法可用于大量数据标注昂贵且困难的场景,例如,在存在强噪声扰动或局部遮挡的地球物理勘探数据中,可利用该方法约束模型以保持对储层属性及热异常预测的一致性 (Wang et al., 2023a; 朱世龙等,2025).图5中展示了基于半监督学习的孔隙度预测流程,通过半监督学习能够充分利用无井、少井区域多源观测数据,提高热储物性参数的预测精度 (Chen et al., 2025b).

2.4 强化学习

强化学习以智能体与环境交互为核心,通过奖赏机制逐步学习最优决策策略,在学习过程中,智能体通过感知所处环境的状态对动作的反应以指导更好的动作,从而最大化可获得的奖励,最早被应用于自动驾驶和游戏AI等场景.强化学习方法可将地热资源勘探开发转化为一个序列决策问题,可用于地热钻井轨迹优化、生产制度调控及储层长期管理等多个方面(Wang et al., 2023bChen et al., 2025a).如图6所示,可通过强化学习方法设计地热勘探智能体,并基于地质、地球物理、地球化学等多源数据构建地热地质环境数值模型.智能体可采取的动作可包括布设新勘探钻孔、调整地球物理测线部署等,可设计奖励函数为成功钻遇热储为正奖励,且对钻井成本提升、储层压力下降过快等不利现象给予负奖励.地热勘探智能体与模拟真实地质情况的数值模拟器不断交互,可学习如何以最小的勘探成本高效圈定地热资源富集区与有利方向.

3 深度学习赋能深层地热资源勘探研究实例

3.1 地热数据智能处理与解译

地热资源勘查的核心目标是识别并评估具有经济开发价值的热储系统,这就要求对地热系统形成与分布的关键地质要素进行精准识别并解译.深度学习技术能够实现自动化高维特征提取与数据增强,具备高效的数据融合与多数据联合解译能力,可有效赋能多源地热数据智能处理与解释,为地热资源靶区优选、资源量评估和开发方案设计提供关键数据基础.

国内外学者成功利用深度学习技术进行多类型地热数据处理与解译,有效服务了区域地热资源有利区预测、地热系统关键要素提取、地热资源远景区预测等(董珮瑶等, 2025; 李月等, 2025).例如,Lösing and Ebbing (2021) 提出了一种基于XGBoost机器学习的大地热流预测方法,该方法在澳大利亚已知地温井区验证的预测结果准确率达到70%以上,并利用最优模型预测了南极洲的大地热流值.Shahdi et al. (2021) 利用美国东北部20 750口井的井底温度和地质信息,利用包括深层神经网络的多种机器学习方法预测了美国东北部地下不同深度的热流和地温梯度.Vesselinov et al. (2022)利用非负矩阵分解和改进的无监督学习聚类算法识别了美国新墨西哥州南部地热数据集中的隐藏地热系统模式,建立了研究区不同区域中水热型地热系统特征的隐含空间关联,并提取了研究区域内中低温水热型地热系统的关键要素,有效根据地质、地球物理、水文地质和地热属性分析了研究区的地热条件和不确定性.Zhou et al. (2024)基于 U⁃Net框架提出了一种结合多元数据约束和加权函数的重力反演方法,并基于遥感重力观测数据识别了共和盆地地热系统内的热源,在此基础上通过物理信息约束的重力反演方法提高了深部地质空间的反演精度和分辨率,揭示了研究区深部地下空间10~35 km内存在的大量低密度体可能是共和盆地地热系统的热源(图7).杨艺等(2024) 结合长短时记忆网络与粒子群优化算法,精准预测了共和盆地恰卜恰地区地热储层温度.Sun et al. (2021) 利用青海省干热岩微震事件数据,基于循环神经网络实现了干热岩资源微震事件的精准识别,可实现干热岩水力压裂过程中的实时监测.

3.2 地热储层智能表征与建模

地热储层是控制地热资源赋存与开发潜力的核心地质单元,其空间分布、热储结构、孔渗特征以及流体热力学状态直接决定了地热资源的可采性与开发效率.传统地热储层精细表征与建模方法往往依赖钻井、地震、电磁、地质、地球化学等多源数据的联合解释,并通过地质建模技术实现储层结构和物性参数的三维重建(吴冲龙和刘刚, 2015; 陈麒玉等, 2020).然而,由于不同类型、不同来源数据往往存在尺度差异,且不同数据之间存在隐含的相关关系,导致传统地热储层不确定性强、建模分辨率低.深度学习技术可直接从多元稀疏地热观测数据中学习到空间耦合特征,并利用深层神经网络建立从多源信息到储层特征之间的非线性映射关系,实现地热储层的高效精细表征与建模.

深度学习技术在储层智能表征与建模领域取得了显著成效 (Feng et al., 2022; 宋随宏等, 2022; Cui et al., 2023; 陈麒玉等, 2025;Chen et al., 2026b).Tut Haklidir and Haklidir(2020)利用土耳其西安纳托利亚地区的83个温泉和地热井的水化学数据,利用深度神经网络准确预测了地热储层温度,该预测方法也适用于预测采样期间二氧化硅沉淀的情况.Cui et al. (2024)提出了一种多条件融合神经网络,可以用于提取多源数据之间的隐含非线性相关关系,并能够实现储层智能高效建模.在MCF⁃Net的基础上,Cui et al. (2025)构建了知识引导神经网络,能够支持多类型观测数据输入并通过深层神经网络直接输出三维储层模型,并在神经网络模型生成过程中嵌入了知识引导机制,以实现知识‒数据联合驱动的裂缝型储层高效建模.Ishitsuka et al. (2025) 利用物理信息神经网络在热场和流场物理约束下精准模拟地热田的温度、压力和渗透率等物性参数(图8).Gao and Zhao (2024) 基于U⁃Net神经网络框架并融合了注意力机制和残差结构,能够根据多种岩石参数和地质构造实现精准的干热岩温度场模拟.Hu et al. (2026) 同样在U⁃Net神经网络框架下结合空间分组增强模块和动态蛇形卷积,能够从花岗岩破坏影像中提取得到裂缝的空间形态和特征,为干热岩储层的裂缝识别与特征提取提供了一种精确有效的方法.

3.3 地热资源智能评价

地热资源评价是地热勘探开发全流程中的核心环节,是评估资源潜力与制定开发方案的关键前提,旨在定量表征地热资源空间分布特征、储量以及可采资源量,从而为地热精细勘探部署与综合开发提供决策.传统地热资源评价主要依赖容积法、热流密度法或数值模拟法进行定量分析,深度学习通过数据驱动的特征学习与模式识别,有望实现在数据稀疏条件下实现高精度、快速、动态地热资源智能评价 (Wang et al., 2024).

Ishitsuka et al. (2021) 结合了贝叶斯、神经网络方法的优势精准预测了日本角田地热田的地下温度分布,预测得到该地热田地下3.7 km处的温度可达500 ℃,该预测结果与基于地质和地球物理约束下的数值模拟结果能够保持高度一致.Wang et al. (2025b) 根据岩性、断层分布、地表水系统、地震点、岩浆岩等多源遥感数据,利用多种深度学习技术评价了杭嘉湖平原的地热资源潜力.Brown et al. (2022) 搭建了一个贝叶斯神经网络并利用美国内华达州研究区对应的10个地质和地球物理特征图来预测地热潜力,通过贝叶斯神经网络能够在考虑到预测不确定性的前提下确定地热潜力最高的区域.Zhang et al. (2023) 以吉林省前郭县地热资源为研究对象,利用机器学习方法确定了地热资源评价过程中影响最大的参数,并利用这些参数进行地热资源评价与储量估算,得出长岭凹陷南部青山口组和泉头组具有良好的地热开发潜力(如图9所示).Cheng et al. (2024) 利用包括反向传播神经网络等6种机器学习和深度学习方法对地热勘探目标区进行了预测并对研究区地热潜力进行了分类评价.

3.4 钻井参数智能预测

地质工程钻探是深层地热资源勘探的验证环节,也是勘探成本最高、风险最大的阶段之一,而深层地热井通常具有高温、高压、低渗透、强非均质等特征,井下地层条件复杂且不确定性大,若钻井过程中参数控制不当极易引发一系列工程风险.深度学习技术能够自动从高维、多噪声的时序监测数据中学习地层特征和钻井相应规律,为钻井参数预测与钻井安全提供关键技术支持和保障 (刘清友和严梁柱, 2025).

近年来,国内外学者通过结合各类机器学习、深度学习方法开展了一系列钻井参数智能预测研究 (李宏波等, 2022).Yehia et al. (2024) 基于美国犹他州FORGE地热储层的钻井数据,利用了10种最先进的机器学习方法预测了钻井机械钻速.Zhai et al. (2025) 同样利用FORGE地热储层钻井数据训练了反向传播神经网络,该神经网络能够有效利用最大钻速、泵压、扭矩和流入量等特征有效预测钻井深度值,预测准确率可达99%以上,能够有效优化钻井性能.Kiran et al. (2022) 利用机器学习方法构建了一个用于预测井漏问题的预测模型,利用地面钻井数据监督深度学习算法的输入,能够识别可能发生井漏问题的危险区域.付加胜等 (2021) 提出利用卷积神经网络和长短时记忆网络相结合的深度学习方法预测钻井溢流问题;首先利用特征筛选得到包括井深、钻头深度、泵冲和立压等16个参数特征,进而通过监测上述特征的时序变化监测钻井漏失的可能性.如图10a所示,CNN⁃LSTM模型在3 307 s预测溢流发生,同时对比图10b所示的录井参数,从第3 868 s开始由于泵冲降低,入口流量同时降低,而出口流量没有降低,后续还略有升高,表明确实发生溢流,验证了深度学习能够比录井监测手段提前10 min预测溢流发生,以保证井筒安全.

4 结论与展望

深层地热资源勘探的核心目标是经济、安全、高效地探明并开发利用埋藏于地壳深处的高品位地热能,涵盖地热靶区筛选、热储特征表征、资源潜力评价、开发风险识别及钻井井位优选等关键环节.基于前述对地热勘探技术体系、深度学习方法及其融合应用的系统综述与典型实践案例分析可知,深度学习技术在多源地热数据解译、热储参数精细建模、钻井参数预测与地热资源评价等方面已展现出显著优势与广阔潜力,有望在深层地热资源勘探的全流程中发挥重要作用 (李根生等, 2024b, 2025; 罗功伟等, 2025).其中,以卷积神经网络为基础的视觉类深度学习模型能够在地震、电磁、红外遥感等地球物理数据中准确提取与地热资源相关的空间特征,助力深层地热靶区自动识别与热异常监测.图神经网络可构建热源、构造等关键要素与地热系统之间的拓扑关联,帮助识别深层地热储层连通性与流体运移通道.生成对抗网络与扩散模型等生成式神经网络能够在观测数据约束下预测热储物性参数.物理信息神经网络能够将热传导、渗流、热力学方程嵌入到损失函数中,能够在保证物理一致性的同时预测深层热储温度场与流体场.循环神经网络、长短期记忆网络等时间序列数据分析方法,能够用于预测地热井温度、压力等动态参数,实现钻井安全监测与动态预警.上述深度学习方法能够从地热数据处理、多源地热数据集成与融合、地热储层建模与预测、地热资源评价、地热资源钻探多个方面提升深层地热勘探效率、准确性和精度,有效助力深层地热资源勘探全流程数字化、智能化建设.然而,面对深部地质空间与地热系统的复杂性、数据稀缺性等问题,深度学习的有效应用仍面临多重挑战.为此,本文进一步从以下三个方面系统阐述深度学习与深层地热勘探交叉领域的关键问题与未来发展路径(图11).

4.1 多模态地热勘探数据高效融合与利用

地热勘探数据具有典型的多源异构特征,涵盖地质、地球物理、地球化学、测井和测温等不同模态的数据形式(如文本、图像、三维数据体等).由于这些数据采集自不同学科的不同方法,导致数据格式、数据描述的空间尺度、数据的物理意义等具有显著差异,但又存在着隐含的数据关联、物理关联和空间耦合关系(Hopcroft and Gallagher, 2023Chao et al., 2024).深层地质空间数据采集难度大,往往缺乏高质量地球物理勘探数据和深地井下测量数据 (Nooshiri et al., 2022Dong et al., 2025).在有限数量和质量的深层地热勘探数据条件下,如何充分挖掘多源信息并实现有效融合和资源预测,是深层地热资源勘探所面临的重要挑战.而深度学习方法往往需要大量高质量的数据来进行神经网络模型训练和参数优化,这在深层地热资源勘探领域难以实际应用.多模态深度学习能够提取、融合、利用不同模态数据中的深层次特征,以弥补单一模态在空间感知与预测精度方面的不足,从而实现对深部地热系统更为全面和精细的认知.因此,借助多模态深度学习技术实现对多源异构地热数据特征的提取、融合与感知,将是提升深层地热勘探精度和效率的有效途径.

4.2 地热领域的可解释与可信人工智能

深度神经网络是包含数百万个参数和多个隐含层的“黑盒”模型,这使人们难以理解每个神经元对最终输出结果的具体影响和贡献.而深层地质空间受多期次、多体制构造演化控制,导致深层地热系统模式异常复杂,深部热储结构、物性参数空间展布等具有强烈的非均质性和空间各向异性 (Tomac and Sauter, 2018Zhang and Zhao, 2020; Ma, 2023).若仅依赖纯数据驱动的深度学习模型,虽能在经验数据分布上实现高精度拟合,但结果往往缺乏可解释性与物理合理性,这种缺乏可解释性和可信度的“黑盒”特征为深层地热工程钻探与开发等安全敏感任务带来严重威胁 (Gunning et al., 2019Kaur et al., 2022).因此,需谨慎验证深度学习方法预测结果,并重点关注地热领域的可解释与可信人工智能技术研发(如领域知识嵌入、多物理场模型约束等),以降低深度学习“黑盒”模型带来的风险,为搭建“人在回路”的智能化、标准化深部地热资源勘探与高效开发工作流程提供关键技术支撑.

4.3 地热领域智能计算基座与大模型

近年来,科学智能计算基座与垂直领域大模型已成为各行业研究热点,也为深层地热勘探技术革新提供了新范式 (Xu et al., 2021Chen et al., 2026a).地热领域科学智能计算基座旨在提供一个地热勘探与开发全流程的数据与算法基础设施,涵盖数据治理、语义建模、知识图谱、智能索引、模型调度与可视化计算环境等多个层级,以实现地热勘探开发全流程多模态地热数据标准化、语义标注与智能索引,能够为地热垂直领域大模型提供高质量领域数据基础和地热领域语义指导(Mudunuru et al., 2022).地热垂直领域大模型则可通过多任务、多模态的跨阶段预训练与多源知识蒸馏,能够在多源数据与物理模拟任务中学习共性规律,并在有限数据样本条件下保持知识迁移与任务泛化能力(Zhao et al., 2024),为克服深层地热勘探数据稀缺性与高成本问题提供可行途径.

上述三个方面层层递进,首先解决有限深层地热资源勘探数据质量下的多源异构数据融合与利用,进而发展领域专家可信的可解释深度学习,再到构建支撑深层地热勘探开发智能决策的核心基础设施,推动地热勘探由“经验驱动”向“知识驱动”再到“智能驱动”的跨越.未来,随着多源地热数据持续积累、算力基础不断提升与可解释理论方法的完善,深度学习将在深层地热资源勘探开发全流程中发挥更深远的科学与工程价值,为我国乃至全球深层地热能规模化开发提供关键技术支撑与理论引领.

参考文献

[1]

Afshar, A., Norouzi, G. H., Moradzadeh, A., 2023. Exploring Geothermal Potential through Multi⁃Modal Geophysical Data Integration: Gravity, Magnetic, and Magnetotelluric Prospecting. International Journal of Mining and Geo⁃Engineering, 57(4): 427-434. https://doi.org/10. 1007/s00024⁃016⁃1448⁃z

[2]

Asghar, S., Choi, J., Yoon, D., et al., 2020. Spatial Pseudo⁃Labeling for Semi⁃Supervised Facies Classification. Journal of Petroleum Science and Engineering, 195: 107834. https://doi.org/10.1016/j.petrol.2020.107834

[3]

Assouline, D., Mohajeri, N., Gudmundsson, A., et al., 2019. A Machine Learning Approach for Mapping the very Shallow Theoretical Geothermal Potential. Geothermal Energy, 7(1): 19. https://doi.org/10.1186/s40517⁃019⁃0135⁃6

[4]

Bénard, B., Famin, V., Agrinier, P., et al., 2024. Use of Cold Waters Geochemistry as a Geothermal Prospecting Tool for Hidden Hydrothermal Systems in Réunion Island. Communications Earth & Environment, 5: 55. https://doi.org/10.1038/s43247⁃024⁃01210⁃3

[5]

Brown, S., Rodi, W. L., Seracini, M., et al., 2022. Bayesian Neural Networks for Geothermal Resource Assessment: Prediction with Uncertainty. arXiv, 2209.15543. https://arxiv.org/abs/2209.15543

[6]

Carlo, C., 2025. Zanskar Validates AI⁃Based Discovery Method with Drilling at Pumpernickel Geothermal Site, Nevada. [2025-9-19] https://www.thinkgeoenergy.com/zanskar⁃validates⁃ai⁃based⁃discovery⁃method⁃with⁃drilling⁃at⁃pumpernickel⁃geothermal⁃site⁃nevada/

[7]

Chang, K., Zhang, Q. J., Jiang, Q. Y., et al., 2025. Research Status of Geothermal Energy Detection Technology in Middle⁃Deep Depths in China. Progress in Geophysics, 40(1): 54-69 (in Chinese with English abstract).

[8]

Chao, J. Q., Zhao, Z. F., Xu, S. G., et al., 2024. Geothermal Target Detection Integrating Multi⁃Source and Multi⁃Temporal Thermal Infrared Data. Ore Geology Reviews, 167: 105991. https://doi.org/10.1016/j.oregeorev.2024.105991

[9]

Chen, C. X., Yan, J. Y., Zhou, W. Y., et al., 2020. Status and Prospects of Geophysical Method Used in Geothermal Exploration. Progress in Geophysics, 35(4): 1223-1231 (in Chinese with English abstract).

[10]

Chen, G. D., Jiao, J. J., Wang, Z. Z., et al., 2025a. Multi⁃Fidelity Machine Learning with Knowledge Transfer Enhances Geothermal Energy System Design and Optimization. Advances in Geo⁃Energy Research, 16(3): 244-259. https://doi.org/10.46690/ager.2025.06.05

[11]

Chen, H. L., Chen, H. Z., Zhao, Z. J., et al., 2026a. An Overview of Domain⁃Specific Foundation Model: Key Technologies, Applications and Challenges. Science China Information Sciences, 69(1): 111301. https://doi.org/10.1007/s11432⁃025⁃4498⁃2

[12]

Chen, M. X., 1992. Advances of Studies of Geothermal Resources in China. Advance in Earth Sciences, 7(3): 9-14 (in Chinese with English abstract).

[13]

Chen, Q. Y., Liu, G., He, Z. W., et al., 2020. Current Situation and Prospect of Structure⁃Attribute Integrated 3D Geological Modeling Technology for Geological Big Data. Bulletin of Geological Science and Technology, 39(4): 51-58 (in Chinese with English abstract).

[14]

Chen, Q. Y., Xun, L., Cui, Z. S., et al., 2025. Recent Progress and Development Trends of Three⁃Dimensional Geological Modeling. Bulletin of Geological Science and Technology, 44(3): 373-387 (in Chinese with English abstract).

[15]

Chen, Q. Y., Zhou, R. H., Chen, D. J., et al., 2026b. A Conditional Masked Autoencoder Network Based on Efficient Multiple⁃Head Self⁃Attention for Characterizing Heterogeneous Reservoirs. Expert Systems with Applications, 296: 128973. https://doi.org/10.1016/j.eswa.2025.128973

[16]

Chen, Y. Y., Zhao, L. X., Liu, J. Y., et al., 2025b. Seismic Porosity Prediction via Semi⁃Supervised Learning: Integrating a Low⁃Frequency Model and a Closed⁃Loop Network Structure. IEEE Transactions on Geoscience and Remote Sensing, 63: 4506617. https://doi.org/10.1109/TGRS.2025.3589022

[17]

Cheng, Q. M., 2021. What Are Mathematical Geosciences and Its Frontiers? Earth Science Frontiers, 28(3): 6-25 (in Chinese with English abstract).

[18]

Cheng, Q. M., 2025. A New Paradigm for Mineral Resource Prediction Based on Human Intelligence⁃Artificial Intelligence Integration. Earth Science Frontiers, 32(4): 1-19 (in Chinese with English abstract).

[19]

Cheng, X. G., Qiao, W., Hu, D. Q., et al., 2024. Quality Analysis of Machine Learning Methods Applied to the Geothermal Potential Assessment: A Case Study. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 46(1): 854-871. https://doi.org/10.1080/15567036.2023.2291451

[20]

Chopra, S., Sharma, R. K., Bredesen, K., et al., 2021. Seismic Characterization of a Triassic⁃Jurassic Deep Geothermal Sandstone Reservoir, Onshore Denmark, Using Unsupervised Machine Learning Techniques. Interpretation, 9(4): T1097-T1106. https://doi.org/10.1190/int⁃2021⁃0091.1

[21]

Cong, T. T., Tan, H. B., Cong, P. X., et al., 2024. Mechanism Governing on Different Hydrochemical Evolution Processes for Two Types of Travertine and Silica Sinters in Shannan, Tibet. Geological Journal of China Universities, 30(6): 646-659 (in Chinese with English abstract).

[22]

Cui, Z. S., Chen, Q. Y., Liu, G., 2023. A Two⁃Stage Downscaling Hydrological Modeling Approach via Convolutional Conditional Neural Process and Geostatistical Bias Correction. Journal of Hydrology, 620: 129498. https://doi.org/10.1016/j.jhydrol.2023.129498

[23]

Cui, Z. S., Chen, Q. Y., Luo, J., et al., 2024. Characterizing Subsurface Structures from Hard and Soft Data with Multiple⁃Condition Fusion Neural Network. Water Resources Research, 60(11): e2024WR038170. https://doi.org/10.1029/2024WR038170

[24]

Cui, Z. S., Jiang, S., Chen, Q. Y., et al., 2025. Characterization of Reservoir Structures with Knowledge⁃ Informed Neural Network. SPE Journal, 30(8): 4469-4486. https://doi.org/10.2118/228279⁃pa

[25]

Dong, B., Li, B., Song, R. C., et al., 2025. Detection and Constraints of Geothermal Latent Heat Zones under the Complex Terrain of the Western Sichuan Plateau: a Fusion of Multi⁃Source Temporal Remote Sensing Data. Geothermics, 130: 103287. https://doi.org/10.1016/j.geothermics.2025.103287

[26]

Dong, P. Y., Du, L., Zhao, L., et al., 2025. Application of Machine Learning Models for Groundwater Temperature Prediction in Geothermal Development. Bulletin of Geological Science and Technology, 44(3): 388-398 (in Chinese with English abstract).

[27]

Dong, Y., Li, G. H., Gao, P. J., et al., 2020. The Application of Fretting Exploration Technology in the Exploration of Middle and Deep Clean Energy. Geophysical and Geochemical Exploration, 44(6): 1345-1351 (in Chinese with English abstract).

[28]

Feng, R. H., Grana, D., Mukerji, T., et al., 2022. Application of Bayesian Generative Adversarial Networks to Geological Facies Modeling. Mathematical Geosciences, 54(5): 831-855. https://doi.org/10.1007/s11004⁃022⁃09994⁃w

[29]

Fu, G. Q., Peng, S. P., Wang, R. Z., et al., 2022. Seismic Prediction and Evaluation Techniques for Hot Dry Rock Exploration and Development. Journal of Geophysics and Engineering, 19(4): 694-705. https://doi.org/10.1093/jge/gxac042

[30]

Fu, J. S., Liu, W., Han, X. S., et al., 2021. CNN⁃LSTM Fusion Network Based Deep Learning Method for Early Prediction of Overflow. China Petroleum Machinery, 49(6): 16-22 (in Chinese with English abstract).

[31]

Gao, W. L., Zhao, J. T., 2024. Deep⁃Time Temperature Field Simulation of Hot Dry Rock: A Deep Learning Method in both Time and Space Dimensions. Geothermics, 119: 102978. https://doi.org/10.1016/j.geothermics.2024.102978

[32]

Gao, Z. J., Sun, Z. J., Yang, Y. H., et al., 2019. Occurrence Characteristics and Hydrochemical Characteristics of Geothermal Water in Shandong Province. Science Technology and Engineering, 19(20): 85-90 (in Chinese with English abstract).

[33]

Gunning, D., Stefik, M., Choi, J., et al., 2019. XAI—Explainable Artificial Intelligence. Science Robotics, 4(37): eaay7120. https://doi.org/10.1126/scirobotics.aay7120

[34]

Guo, J. T., Xu, X. C., Wang, L. Y., et al., 2024. GeoPDNN 1.0: A Semi⁃Supervised Deep Learning Neural Network Using Pseudo⁃Labels for Three⁃Dimensional Shallow Strata Modelling and Uncertainty Analysis in Urban Areas from Borehole Data. Geoscientific Model Development, 17(3): 957-973. https://doi.org/10.5194/gmd⁃17⁃957⁃2024

[35]

Guo, Q. H., 2020. Magma⁃Heated Geothermal Systems and Hydrogeochemical Evidence of Their Occurrence. Acta Geologica Sinica, 94(12): 3544-3554 (in Chinese with English abstract).

[36]

He, J. S., 2020. New Research Progress in Theory and Application of Wide Field Electromagnetic Method. Geophysical and Geochemical Exploration, 44(5): 985-990 (in Chinese with English abstract).

[37]

Hopcroft, P. O., Gallagher, K., 2023. Global Variability in Multi⁃Century Ground Warming Inferred from Geothermal Data. Geophysical Research Letters, 50(13): e2023GL104631. https://doi.org/10.1029/2023GL104631

[38]

Hu, M. G., Fu, T. Y., Yang, X. B., et al., 2026. Fracture Identification in Hot Dry Rock Using TSD⁃Unet: From Feature Extraction to Quantitative Analysis of Geometric Parameters. Geothermics, 136: 103569. https://doi.org/10.1016/j.geothermics.2025.103569

[39]

Huang, X. L., Han, Y. J., Xiao, Z. C., et al., 2023. Fluoride Occurrence in Geothermal Water of Fault Zone Area, Southeast China. Chemosphere, 328: 138468. https://doi.org/10.1016/j.chemosphere.2023.138468

[40]

Huang, X. L., Wang, S., Wang, H., et al., 2024. Hidden Magma Reservoirs: Insights from Helium and Carbon Isotopic Compositions in Shallow Groundwater Beneath the Wudalianchi Volcanic Field, Northeastern China. Chemical Geology, 666: 122310. https://doi.org/10.1016/j.chemgeo.2024.122310

[41]

Huo, C., Lin, Y. T., Li, G., et al., 2023. China’s Geothermal Resource Exploration Technology Research Progress under the Background of Carbon Neutrality. Science Technology and Engineering, 23(12): 4917-4927 (in Chinese with English abstract).

[42]

Ishitsuka, K., Ishizu, K., Watanabe, N., et al., 2025. Reliable and Practical Inverse Modeling of Natural⁃State Geothermal Systems Using Physics⁃Informed Neural Networks: Three⁃Dimensional Model Construction and Assimilation with Magnetotelluric Data. Journal of Geophysical Research: Machine Learning and Computation, 2(3): e2025JH000683. https://doi.org/10.1029/2025JH000683

[43]

Ishitsuka, K., Kobayashi, Y., Watanabe, N., et al., 2021. Bayesian and Neural Network Approaches to Estimate Deep Temperature Distribution for Assessing a Supercritical Geothermal System: Evaluation Using a Numerical Model. Natural Resources Research, 30(5): 3289-3314. https://doi.org/10.1007/s11053⁃021⁃09874⁃w

[44]

Ishizu, K., Ogawa, Y., Nunohara, K., et al., 2022. Estimation of Spatial Distribution and Fluid Fraction of a Potential Supercritical Geothermal Reservoir by Magnetotelluric Data: A Case Study from Yuzawa Geothermal Field, NE Japan. Journal of Geophysical Research: Solid Earth, 127(2): e2021JB022911. https://doi.org/10.1029/2021JB022911

[45]

Jiang, S., Wang, S., Qi, S. H., et al., 2020. Recent Advances in the Data⁃Driven Play Fairway Analysis for Geothermal Exploration. Geological Journal of China Universities, 26(1): 111-120 (in Chinese with English abstract).

[46]

Kaur, D., Uslu, S., Rittichier, K. J., et al., 2022. Trustworthy Artificial Intelligence: A Review. ACM Computing Surveys, 55(2): 1-38. https://doi.org/10.1145/3491209

[47]

Kiran, R., Dansena, P., Salehi, S., et al., 2022. Application of Machine Learning and Well Log Attributes in Geothermal Drilling. Geothermics, 101: 102355. https://doi.org/10.1016/j.geothermics.2022.102355

[48]

Kristensen, L., Hjuler, M. L., Frykman, P., et al., 2016. Pre⁃Drilling Assessments of Average Porosity and Permeability in the Geothermal Reservoirs of the Danish Area. Geothermal Energy, 4(1): 6. https://doi.org/10.1186/s40517⁃016⁃0048⁃6

[49]

Kuang, L. C., Liu, H., Ren, Y. L., et al., 2021. Application and Development Trend of Artificial Intelligence in Petroleum Exploration and Development. Petroleum Exploration and Development, 48(1): 1-11 (in Chinese with English abstract).

[50]

LeCun, Y., Bengio, Y., Hinton, G., 2015. Deep Learning. Nature, 521(7553): 436-444. https://doi.org/10.1038/nature14539

[51]

Lei, Y. D., Yuan, Y. J., Qin, G. X., et al., 2023. Analysis of Thermal Storage Characteristics of the Guide Zhacang Geothermal Field in Gonghe Basin Based on Logging Data. Acta Geoscientica Sinica, 44(1): 145-157 (in Chinese with English abstract).

[52]

Lei, Y. K., Wang, S., Huang, X. L., et al., 2025. Hydrochemical Characteristics and Formation⁃Evolution Analysis of Medium⁃Low Temperature Geothermal Systems with High Salinity in Coastal Western Guangdong. Earth Science, 50(9): 3616-3630 (in Chinese with English abstract).

[53]

Li, C., Zhang, P., Dai, L., et al., 2021. Research on the Application of Comprehensive Geophysical Prospecting in Middle⁃Deep Geothermal Exploration. Progress in Geophysics, 36(2): 611-617 (in Chinese with English abstract).

[54]

Li, G. S., Song, X. Z., Shi, Y., et al., 2024a. Current Status and Construction Scheme of Smart Geothermal Field Technology. Petroleum Exploration and Development, 51(4): 899-909 (in Chinese with English abstract).

[55]

Li, G. S., Song, X. Z., Shi, Y., et al., 2024b. Current Status and Development Trends of Smart Geothermal Field Technology. Petroleum Exploration and Development, Online (in Chinese with English abstract). https://link.cnki.net/urlid/11.2360.TE.20240619.2200.004

[56]

Li, G. S., Wang, T. Y., Li, J., et al., 2025. Pathways and Prospects for Intelligent and Green Development of Oil and Gas Driven by Multi⁃Energy Integration. Xinjiang Oil & Gas, 21(3): 1-13 (in Chinese with English abstract).

[57]

Li, H. B., Luo, P. Y., Bai, Y., et al., 2022. Summary for Machine Learning Algorithms and Their Applications in Drilling Engineering. Xinjiang Oil & Gas, 18(1): 1-13 (in Chinese with English abstract).

[58]

Li, J. T., Wu, X. M., Ye, Y. M., et al., 2023. Unsupervised Contrastive Learning for Seismic Facies Characterization. Geophysics, 88(1): WA81⁃WA89. https://doi.org/10.1190/geo2022⁃0148.1

[59]

Li, Y., Dong, M., Li, M., 2025. Research Advances in the Application of Machine Learning to Geothermal Heat Flow. Progress in Geophysics, 40(6): 2460-2475 (in Chinese with English abstract).

[60]

Liu, Q. Y., Yan, L. Z., 2025. Current Status and Progress of Research on Intelligent Drill Bits. Acta Petrolei Sinica, 46(6): 1193-1202 (in Chinese with English abstract).

[61]

Liu, Q., Lin, T. Y., Yang, M., et al., 2022. Micropore Structure and Physical Property of Geothermal Reservoir of Wumishan Formation in Beijing Area. Geological Bulletin of China, 41(4): 657-668 (in Chinese with English abstract).

[62]

Liu, Y. C., Liu, K., Sun, Y., et al., 2015. Geochemical Characteristics of Geothermal Water in Beijing. South⁃to⁃North Water Transfers and Water Science & Technology, 13(2): 324-329 (in Chinese with English abstract).

[63]

Lösing, M., Ebbing, J., 2021. Predicting Geothermal Heat Flow in Antarctica with a Machine Learning Approach. Journal of Geophysical Research: Solid Earth, 126(6): e2020JB021499. https://doi.org/10.1029/2020JB021499

[64]

Luo, G. W., An, X. P., Yao, W. H., et al., 2025. Application Status and Development Trends of Artificial Intelligence in Logging Interpretation for Unconventional Oil and Gas Reservoirs. Petroleum Science Bulletin, 10(5): 908-925 (in Chinese with English abstract).

[65]

Lyu, Q., Peng, P., Ji, X. J., et al., 2025. Application of 3D Inversion of Controlled Source Audio⁃Frequency Magnetotelluric in Deep Geothermal Structure Exploration—A Case Study in Volcanic Region of Eastern Zhejiang. Chinese Journal of Engineering Geophysics, 22(3): 416-425 (in Chinese with English abstract).

[66]

Ma, Y. S., 2023. Deep Geothermal Resources in China: Potential, Distribution, Exploitation, and Utilization. Energy Geoscience, 4(4): 100209. https://doi.org/10.1016/j.engeos.2023.100209

[67]

Moradi, M., Basiri, S., Kananian, A., et al., 2015. Fuzzy Logic Modeling for Hydrothermal Gold Mineralization Mapping Using Geochemical, Geological, ASTER Imageries and Other Geo⁃Data, a Case Study in Central Alborz, Iran. Earth Science Informatics, 8(1): 197-205. https://doi.org/10.1007/s12145⁃014⁃0151⁃9

[68]

Mudunuru, M. K., Vesselinov, V. V., Ahmmed, B., 2022. Geothermalcloud: Machine Learning for Geothermal Resource Exploration. Journal of Machine Learning for Modeling and Computing, 3(4): 57-72. https://doi.org/10.1615/jmachlearnmodelcomput.2022046445

[69]

Nooshiri, N., Bean, C. J., Dahm, T., et al., 2022. A Multibranch, Multitarget Neural Network for Rapid Point⁃Source Inversion in a Microseismic Environment: Examples from the Hengill Geothermal Field, Iceland. Geophysical Journal International, 229(1): 999-1016. https://doi.org/10.1093/gji/ggab511

[70]

Palo, M., Ogliari, E., Sakwa, M., 2024. Spatial Pattern of the Seismicity Induced by Geothermal Operations at the Geysers (California) Inferred by Unsupervised Machine Learning. IEEE Transactions on Geoscience and Remote Sensing, 62: 5905813. https://doi.org/10.1109/TGRS.2024.3361169

[71]

Pang, Z. H., Luo, J., Cheng, Y. Z., et al., 2020. Evaluation of Geological Conditions for the Development of Deep Geothermal Energy in China. Earth Science Frontiers, 27(1): 134-151 (in Chinese with English abstract).

[72]

Piana Agostinetti, N., Licciardi, A., Piccinini, D., et al., 2017. Discovering Geothermal Supercritical Fluids: A New Frontier for Seismic Exploration. Scientific Reports, 7: 14592. https://doi.org/10.1038/s41598⁃017⁃15118⁃w

[73]

Qin, T. C., Deng, J. Z., Chen, H., et al., 2020. Application of Audio Frequency Magnetotelluric Method Three⁃Dimensional Inversion in Geothermal Exploration in Fengcheng of Jiangxi, China. Science Technology and Engineering, 20(14): 5489-5498 (in Chinese with English abstract).

[74]

Qiao, Y., Zhang, H., 2025. Methodology and Application of Deep Geothermal Sounding in Low⁃Resistance Cover Areas. Applied Geophysics, 22(1): 99-109. https://doi.org/10.1007/s11770⁃023⁃1026⁃y

[75]

Ragueh, R. R., Tarits, P., Hautot, S., et al., 2024. Inversion of Gravity Data Constrained by a Magnetotelluric Resistivity Model: Application to the Asal Rift, Djibouti. Journal of Geophysical Research: Solid Earth, 129(8): e2023JB028484. https://doi.org/10.1029/2023JB028484

[76]

Rangel⁃Arista, J. A., Zarrouk, S. J., Kaya, E., et al., 2025. Downflows during Transient Geothermal Well Test Analysis. Geothermics, 125: 103158.

[77]

Salimzadeh, S., Grandahl, M., Medetbekova, M., et al., 2019. A Novel Radial Jet Drilling Stimulation Technique for Enhancing Heat Recovery from Fractured Geothermal Reservoirs. Renewable Energy, 139: 395-409. https://doi.org/10.1016/j.renene.2019.02.073

[78]

Shahdi, A., Lee, S., Karpatne, A., et al., 2021. Exploratory Analysis of Machine Learning Methods in Predicting Subsurface Temperature and Geothermal Gradient of Northeastern United States. Geothermal Energy, 9(1): 18. https://doi.org/10.1186/s40517⁃021⁃00200⁃4

[79]

Shebl, A., Abdellatif, M., Badawi, M., et al., 2023. Towards Better Delineation of Hydrothermal Alterations via Multi⁃Sensor Remote Sensing and Airborne Geophysical Data. Scientific Reports, 13: 7406. https://doi.org/10.1038/s41598⁃023⁃34531⁃y

[80]

Song, S. H., Shi, Y. Q., Hou, J. G., 2022. Review of a Generative Adversarial Networks(GANs)⁃Based Geomodelling Method. Petroleum Science Bulletin, 7(1): 34-49 (in Chinese with English abstract).

[81]

Sun, F., Hu, H. T., Zhao, F., et al., 2021. Micro⁃Seismic Event Detection of Hot Dry Rock Based on the Gated Recurrent Unit Model and a Support Vector Machine. Acta Geologica Sinica (English Edition), 95(6): 1940-1947. https://doi.org/10.1111/1755⁃6724.14882

[82]

Sun, H. Q., Mao, X., Wu, C., et al., 2024. Geothermal Resources Exploration and Development Technology: Current Status and Development Directions. Earth Science Frontiers, 31(1): 400-411 (in Chinese with English abstract).

[83]

Sun, X. Y., Zhan, Y., Zhao, L. Q., et al., 2023. Does a Shallow Magma Reservoir Exist in the Wudalianchi Volcanic Field? Constraints from Magnetotelluric Imaging. Geophysical Research Letters, 50(17): e2023GL104318. https://doi.org/10.1029/2023GL104318

[84]

Sun, Z. X., Gao, B., Liu, J. H., 2004. Geothermal Gas Geochemistry of the Hengjinghot Springs Area in Jiangxi Province. Geoscience, 18(1): 116-120 (in Chinese with English abstract).

[85]

Tian, B. Q., Pang, Z. H., Kong, Y. L., et al., 2020. Exploration and Amount Fine Evaluation of Geothermal Resources Based on Microtremor Survey Method. Science & Technology for Development, 16(S1): 367-374 (in Chinese with English abstract).

[86]

Tian, J., Pang, Z. H., Li, Y. M., et al., 2022. Research Progress on Geothermal Gas. Acta Geologica Sinica, 96(5): 1752-1766 (in Chinese with English abstract).

[87]

Tomac, I., Sauter, M., 2018. A Review on Challenges in the Assessment of Geomechanical Rock Performance for Deep Geothermal Reservoir Development. Renewable and Sustainable Energy Reviews, 82: 3972-3980. https://doi.org/10.1016/j.rser.2017.10.076

[88]

Tut Haklidir, F. S., Haklidir, M., 2020. Prediction of Reservoir Temperatures Using Hydrogeochemical Data, Western Anatolia Geothermal Systems (Turkey): A Machine Learning Approach. Natural Resources Research, 29(4): 2333-2346. https://doi.org/10.1007/s11053⁃019⁃09596⁃0

[89]

van der Meer, F., Hecker, C., van Ruitenbeek, F., et al., 2014. Geologic Remote Sensing for Geothermal Exploration: A Review. International Journal of Applied Earth Observation and Geoinformation, 33: 255-269. https://doi.org/10.1016/j.jag.2014.05.007

[90]

Van Noorden, R., Perkel, J. M., 2023. AI and Science: What 1 600 Researchers Think. Nature, 621(7980): 672-675. https://doi.org/10.1038/d41586⁃023⁃02980⁃0

[91]

Vesselinov, V. V., Ahmmed, B., Mudunuru, M. K., et al., 2022. Discovering Hidden Geothermal Signatures Using Non⁃Negative Matrix Factorization with Customized K⁃Means Clustering. Geothermics, 106: 102576. https://doi.org/10.1016/j.geothermics.2022.102576

[92]

Wang, D. F., Cheng, J. F., 2023. China’s Dry⁃Hot Rock Resources Present Situation and Prospect of Exploration Technology under Carbon Peaking and Carbon Neutrality Background. Coal Geology of China, 35(4): 43-49 (in Chinese with English abstract).

[93]

Wang, G. F., Ren, H. W., Zhao, G. R., et al., 2021. Analysis and Countermeasures of Ten’pain Points’ of Intelligent Coal Mine. Industry and Mine Automation, 47(6): 1-11 (in Chinese with English abstract).

[94]

Wang, G. L., Lin, W. J., Liu, F., et al., 2023. Theory and Survey Practice of Deep Heat Accumulation in Geothermal System and Exploration Practice. Acta Geologica Sinica, 97(3): 639-660 (in Chinese with English abstract).

[95]

Wang, G. L., Liu, Y. G., Zhu, X., et al., 2020. The Status and Development Trend of Geothermal Resources in China. Earth Science Frontiers, 27(1): 1-9 (in Chinese with English abstract).

[96]

Wang, G. L., Lu, C., 2022. Progress of Geothermal Resources Exploitation and Utilization Technology Driven by Carbon Neutralization Target. Geology and Resources, 31(3): 412-425, 341 (in Chinese with English abstract).

[97]

Wang, L., Yu, Z. W., Zhang, Y. J., et al., 2023a. Review of Machine Learning Methods Applied to Enhanced Geothermal Systems. Environmental Earth Sciences, 82(3): 69. https://doi.org/10.1007/s12665⁃023⁃10749⁃x

[98]

Wang, N. Z., Chang, H. B., Kong, X. Z., et al., 2023b. Deep Learning Based Closed⁃Loop Well Control Optimization of Geothermal Reservoir with Uncertain Permeability. Renewable Energy, 211: 379-394. https://doi.org/10.1016/j.renene.2023.04.088

[99]

Wang, S. Z., Zhou, Y. Q., Zhang, X., et al., 2024. Mapping and Resource Evaluation of Deep High⁃Temperature Geothermal Resources in the Jiyang Depression, China. Energy Geoscience, 5(4): 100320. https://doi.org/10.1016/j.engeos.2024.100320

[100]

Wang, S., Huang, X. H., Han, W., et al., 2023c. Lithological Mapping of Geological Remote Sensing via Adversarial Semi⁃Supervised Segmentation Network. International Journal of Applied Earth Observation and Geoinformation, 125: 103536. https://doi.org/10.1016/j.jag.2023.103536

[101]

Wang, S., Qi, S. H., Huang, X. L., et al., 2025a. Helium Isotopes in Hot Springs of the Karakorum Fault and the Central Pamir: Tracing Mantle Contributions and Tectonic Dynamics. Global and Planetary Change, 253: 104897. https://doi.org/10.1016/j.gloplacha.2025.104897

[102]

Wang, T. Y., Li, G. S., Song, X. Z., et al., 2024. Development of Smart Oil and Gas Fields with Multi⁃Energy Synergy of Wind, Solar, Geothermal, and Energy Storage. Strategic Study of CAE, 26(4): 259-270 (in Chinese with English abstract).

[103]

Wang, Y. H., Zhang, X., Qian, J. F., et al., 2025b. Machine and Deep Learning⁃Based Prediction of Potential Geothermal Areas in Hangjiahu Plain by Integrating Remote Sensing Data and GIS. Energy, 315: 134370.

[104]

Wei, X., Shi, H. J., Chen, S., et al., 2024. Application of Hydrogeochemical Methods in Geothermal Resource Exploration: A Case Study of Yingcheng City, Hubei Province. Bulletin of Geological Science and Technology, 43(3): 68-80 (in Chinese with English abstract).

[105]

Wu, C. L., Liu, G., 2015. Current Situation, Existent Problems, Trend and Strategy of the Construction of “Glass Earth”. Geological Bulletin of China, 34(7): 1280-1287 (in Chinese with English abstract).

[106]

Xin, H., Ning, Y. W., Ping, J. H., et al., 2024. Hydrogeochemical Characteristics and Genesis of Geothermal Fields in the Henan Section of the Neihuang Uplift. Science Technology and Engineering, 24(34): 14537-14550 (in Chinese with English abstract).

[107]

Xin, L., Liu, X. X., Zhang, B., 2021. Land Surface Temperature Retrieval and Geothermal Resources Prediction by Remote Sensing Image: A Case Study in the Shijiazhuang Area, Hebei Province. Journal of Geomechanics, 27(1): 40-51 (in Chinese with English abstract).

[108]

Xu, Y. J., Liu, X., Cao, X., et al., 2021. Artificial Intelligence: A Powerful Paradigm for Scientific Research. Innovation, 2(4): 100179. https://doi.org/10.1016/j.xinn.2021.100179

[109]

Yamaya, Y., 2025. Electromagnetic Exploration of Supercritical/Super⁃Hot Geothermal Systems. Surveys in Geophysics, Online. https://doi.org/10.1007/s10712⁃025⁃09907⁃6

[110]

Yang, Y., Zhao, J. T., Fu, G. Q., 2024. Predicting Geothermal Reservoir Temperature Based on the PSO⁃LSTM Model. Journal of Mining Science and Technology, 9(4): 538-548 (in Chinese with English abstract).

[111]

Yehia, T., Gasser, M., Ebaid, H., et al., 2024. Comparative Analysis of Machine Learning Techniques for Predicting Drilling Rate of Penetration (ROP) in Geothermal Wells: a Case Study of FORGE Site. Geothermics, 121: 103028. https://doi.org/10.1016/j.geothermics.2024.103028

[112]

Yin, R. G., Wei, S., Li, H., et al., 2016. Introduction of Unsupervised Learning Methods in Deep Learning. Computer Systems & Applications, 25(8): 1-7 (in Chinese with English abstract).

[113]

Yue, W. X., He, X., Lai, F., et al., 2024. Assessing the Contribution of Sedimentary Layer Heat Generation Rate to Geothermal Energy Based on Natural Gamma Logging in the Moxi Area of the Sichuan Basin. Mineralogy and Petrology, 44(4): 193-205 (in Chinese with English abstract).

[114]

Yun, X. R., Feng, J. Y., Zheng, H. R., et al., 2026. Hercynian Granite over Cretaceous Sediments Discovered in Northern Hainan Island—New Evidence of Late Mesozoic Extension—Compressional Tectonic Transformation in South China. Geological Review, 72(1): 50-66 (in Chinese with English abstract).

[115]

Zhai, W., Feng, B., Liu, S. Z., et al., 2025. A Shallow Machine Learning Method Based on Geothermal Drilling Data: A Case Study of Well 58⁃32 at the U.S. FORGE Site. Geothermics, 127: 103239. https://doi.org/10.1016/j.geothermics.2024.103239

[116]

Zhang, G. Y., Lin, C. Y., Wang, Z. Z., et al., 2024. Hybrid Knowledge⁃Driven and Data⁃Driven Intelligent Reservoir Characterization and Its Research Progress. Progress in Geophysics, 39(1): 119-140 (in chinese).

[117]

Zhang, J., Xiao, C. L., Yang, W. F., et al., 2023. Probabilistic Geothermal Resources Assessment Using Machine Learning: Bayesian Correction Framework Based on Gaussian Process Regression. Geothermics, 114: 102787. https://doi.org/10.1016/j.geothermics.2023.102787

[118]

Zhang, S. Q., Fu, L., Zhang, Y., et al., 2020. Delineation of Hot Dry Rock Exploration Target Area in the Gonghe Basin Based on High⁃Precision Aeromagnetic Data. Natural Gas Industry, 40(9): 156-169 (in Chinese with English abstract).

[119]

Zhang, Y. L., Zhao, G. F., 2020. A Global Review of Deep Geothermal Energy Exploration: From a View of Rock Mechanics and Engineering. Geomechanics and Geophysics for Geo⁃Energy and Geo⁃Resources, 6(1): 4. https://doi.org/10.1007/s40948⁃019⁃00126⁃z

[120]

Zhao, B. Y., Wang, S., Chen, F., et al., 2024. Hydrogeochemical Characteristics and Genesis of Medium⁃High Temperature Geothermal System in Northeast Margin of Pamir Plateau. Earth Science, 49(10): 3736-3748 (in Chinese with English abstract).

[121]

Zhao, T. J., Wang, S., Ouyang, C. J., et al., 2024. Artificial Intelligence for Geoscience: Progress, Challenges, and Perspectives. The Innovation, 5(5): 100691. https://doi.org/10.1016/j.xinn.2024.100691

[122]

Zheng, S. F., Li, X., Wang, M., 2025. Multivariate Prediction Model of Geothermal Parameters Based on Machine Learning. Energy, 316: 134497. https://doi.org/10.1016/j.energy.2025.134497

[123]

Zhou, F. Y., Jin, L. P., Dong, J., 2017. Review of Convolutional Neural Network. Chinese Journal of Computers, 40(6): 1229-1251 (in Chinese with English abstract).

[124]

Zhou, S., Wei, Y., Lu, P. Y., et al., 2024. Deep⁃Learning Gravity Inversion Method with Depth⁃Weighting Constraints and Its Application in Geothermal Exploration. Remote Sensing, 16(23): 4467. https://doi.org/10.3390/rs16234467

[125]

Zhou, T. Y., Wang, Z. H., Qin, H. Y., et al., 2020. Remote Sensing Extraction of Geothermal Anomaly Based on Terrain Effect Correction. Journal of Remote Sensing, 24(3): 265-276 (in Chinese with English abstract).

[126]

Zhou, Y., Li, S. J., Fan, M., 2011. Study on Sealing Capacity of Cap Rock in the Process of Tectonic Deformation. Chinese Journal of Geology (Scientia Geologica Sinica), 46(1): 226-232 (in Chinese with English abstract).

[127]

Zhou, Y. Z., Zuo, R. G., Liu, G., et al., 2021. The Great⁃Leap⁃Forward Development of Mathematical Geoscience during 2010⁃2019: Big Data and Artificial Intelligence Algorithm Are Changing Mathematical Geoscience. Bulletin of Mineralogy, Petrology and Geochemistry, 40(3): 556-573, 777 (in Chinese with English abstract).

[128]

Zhu, J., Jin, S., Yang, Y., et al., 2022. Geothermal Resource Exploration in Magmatic Rock Areas Using a Comprehensive Geophysical Method. Geofluids, 2022(1): 5929324. https://doi.org/10.1155/2022/5929324

[129]

Zhu, S. L., Sun, L. X., Zhu, J. B., et al., 2025. An Identification Method for a 3D Geological Anomalous Body Based on a Small Number of Profile Labels. Chinese Journal of Geophysics, 68(3): 1116-1129 (in Chinese with English abstract).

基金资助

国家自然科学基金项目(42502294)

国家重点研发计划(2022YFF08012002)

湖北省国际科技合作项目(2024EHA026)

湖北省自然科学基金项目(2025AFB179)

国家资助博士后研究人员计划(GZB20250110)

湖北省博士后创新人才培养项目(2025HBBSHCXB088)

中央高校基本科研业务费专项资金资助项目(143⁃G1323525070)

湖北省中央引导地方科技发展专项资金项目(2025CSA020)

AI Summary AI Mindmap
PDF (3327KB)

0

访问

0

被引

详细

导航
相关文章

AI思维导图

/