CarbGeo-3D: A Synthetic Three-dimensional Lithology-Resistivity Dataset of Carbonate Geothermal Reservoir
CarbGeo-3D: 炭酸塩岩地熱貯留層のための合成3次元岩相-比抵抗データセット (AI 翻訳)
Zhesi Cui, Qiyu Chen
🤖 gxceed AI 要約
日本語
本データセットは、炭酸塩岩地熱貯留層のための5,000組の3次元岩相モデルと比抵抗ボリュームを提供する。機械学習や物理探査逆問題への応用を目的とし、オープンソース（CC BY 4.0）で公開される。パラメータはユーザーが調整可能であり、地熱貯留層の特性評価の再現性とベンチマークを支援する。
English
CarbGeo-3D is a large-scale synthetic dataset of paired 3D lithological and resistivity models for carbonate geothermal reservoirs. With 5,000 data pairs, it supports data-driven subsurface characterization, machine learning, and geophysical inversion. The dataset is released in multiple formats under CC BY 4.0, and parameters can be adjusted for specific geological settings. This enables reproducible evaluation and benchmarking of geothermal reservoir modeling methods.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本は地熱資源が豊富だが開発が遅れており、本データセットは機械学習による貯留層特性評価の高度化に寄与し、GX政策における地熱の位置づけを強化する可能性がある。
In the global GX context
Geothermal energy is a crucial baseload renewable source globally. This dataset fills a gap in open subsurface data for machine learning, accelerating data-driven reservoir characterization and supporting the energy transition.
👥 読者別の含意
🔬研究者:Useful for benchmarking ML models for subsurface characterization and inversion in geothermal reservoirs.
🏢実務担当者:Can be used to train inversion models for geothermal reservoir exploration, potentially reducing drilling risk.
📄 Abstract（原文）
We introduce CarbGeo-3D, a comprehensive synthetic dataset comprising 5,000 paired 3D lithological models and continuous resistivity volumes. To our knowledge, CarbGeo-3D represents one of the first openly available large-scale datasets that jointly provide paired three-dimensional lithofacies and resistivity models for carbonate geothermal reservoirs while explicitly incorporating geological process knowledge into the generation workflow. CarbGeo-3D is designed to support the development, benchmarking, and reproducible evaluation of data-driven subsurface characterization methods by providing a large collection of geologically plausible carbonate geothermal reservoir models. The synthesized lithology and resistivity models of carbonate geothermal reservoirs are fully aligned in spatial grid size (default 128 × 128 × 128), allowing them to be directly used as paired geological-geophysical data for machine learning, geophysical inversion, uncertainty quantification, and surrogate modeling applications. The dataset is released in several standardized formats. The dataset includes *.npy to be compatible with mainstream deep learning frameworks, *.vtk to support professional geophysical visualization software such as ParaView, and *.gslib to be compatible with geostatistical software packages. The parameters of generating lithology-resistivity data pairs can be flexibly adjusted by the user according to the specific geological and geophysical characteristics of the study area. This dataset is licensed under the CC BY 4.0 open-source license, supporting open sharing and community reuse.
🔗 Provenance — このレコードを発見したソース
- scidb https://doi.org/10.57760/sciencedb.40807first seen 2026-07-29 10:54:14
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。