IN037
Rise of Large Language Models in the Geosciences
Session ID#: 251518
Session Description:
Large language models (LLMs) are rapidly becoming integral tools for geoscientists, enabling new approaches to literature review, data interpretation, simulation support, and cross-domain synthesis. This session invites researchers applying LLMs, particularly domain-specific systems like GeoGPT, to address practical challenges across Earth and environmental sciences.Presentations will explore how LLMs are being used in real-world applications such as taxonomy, data curation, geologic mapping, environmental protection, resource exploration, hazard forecasting and disaster response. We also welcome discussions on interdisciplinary workflows, model evaluation in scientific contexts, ethical and governance issues, and the responsible deployment of generative AI across geoscientific domains.By highlighting cases of applied usage and lessons learned, this session aims to foster collaboration between AI researchers and domain scientists and to present models of best practices for integrating LLMs into geoscientific research and decision-making.
Co-Sponsor(s):
Index Terms:
1910 Data assimilation, integration and fusion [INFORMATICS]
1916 Data and information discovery [INFORMATICS]
1942 Machine learning [INFORMATICS]
1954 Natural language processing [INFORMATICS]
Primary Convener: Grant Boquet, Zhejiang Laboratory, Hangzhou City, China
Conveners: James G Ogg, International Union of Geological Sciences, Deep-time Digital Earth, DDE Research Center of Excellence (Suzhou), Suzhou, China and Jieping Ye, Zhejiang Laboratory, Hangzhou City, China
GIS-Task-Graph: A Case-Based Knowledge Graph to Enhance LLM-Based Agents in Geospatial Analysis (1903626)
Xiang Zhao1, Jie Feng2, Chao Wang3, Linshu Hu4, Sensen Wu1 and Zhenhong Du1, (1)Zhejiang University, Hangzhou, China, (2)Zhejiang University, HangZhou, China, (3)Zhejinag University, HangZhou, China, (4)School of Earth Sciences, Zhejiang University, Hangzhou, Zhejiang Province, China
AI Tools Enabled an Online Database of the ca. 50,000 Fossil Genera of the Treatise of Invertebrate Paleontology (1916954)
James G Ogg1, Aditya Sivathanu2, Kevin Chang2, Wen Du3, Aaron Ault2, Jieping Ye4, Zongyuan Xiang5, Juye Wei5, Hangyang Chen4 and Yufei Ye5, (1)Chengdu University of Technology, Key Laboratory of Deep-time Geography and Environment Reconstruction and Applications, MNR & Institute of Sedimentary Geology, Chengdu, China, (2)Purdue University, Electrical and Computer Engineering, West Lafayette, United States, (3)Geologic Timescale Foundation, West Lafayette, United States, (4)Zhejiang Laboratory, Hangzhou City, China, (5)Zhejiang Lab, GeoGPT, Hangzhou, China
A Large Language Model-Based Python Pipeline for Automated Extraction of Hydroelectric Facility Operational Information (1933037)
Faisal Ashraf1, Hong-Jun Yoon2, Carly Hansen3, Thomas Alan Ruggles4, Debjani Singh3, Esther S Parish3 and Shih-Chieh Kao5, (1)Oak Ridge National Laboratory, Environmental Sciences Division, Oak Ridge TN, United States, (2)Oak Ridge National Laboratory, Computational Sciences and Engineering Division, Oak Ridge, United States, (3)Oak Ridge National Laboratory, Environmental Sciences Division, Oak Ridge, TN, United States, (4)Oak Ridge National Laboratory, Environmental Sciences Division, Oak Ridge, United States, (5)Oak Ridge National Laboratory, Environmental Science Division, Oak Ridge, United States
Transforming Climate Services with LLMs and Multi-Source Data Integration (1943125)
Ivan Kuznetsov1, Antonia Anna Jost1, Dmitrii Pantiukhin2, Boris Shapkin1, Thomas Jung3,4 and Nikolay Koldunov1, (1)Alfred Wegener Institute Helmholtz-Center for Polar and Marine Research Bremerhaven, Climate Dynamics, Bremerhaven, Germany, (2)Alfred Wegener Institute Helmholtz-Center for Polar and Marine Research Bremerhaven, Bremerhaven, Germany, (3)Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research, Bremerhaven, Germany, (4)University of Bremen, Institute of Environmental Physics, Bremen, Germany
RoGO: Fast and Interpretable Graph-Based Reconstruction of Composite Fossil Sequences (1954053)
Shuhan Zhang1, Jiacheng Len2, Grant Boquet3, Hangyang Chen4, Jieping Ye4 and Shu-zhong Shen1, (1)Nanjing University, Nanjing, China, (2)Zhejiang Lab, Hangzhou, China, (3)Organization Not Listed, Washington, United States, (4)Zhejiang Laboratory, Hangzhou City, China
Recent progress of knowledge-infused and AI-driven extension to the Mindat open data service (1969959)
Xiaogang Ma1, Jiyin Zhang1, Weilin Chen2 and Jolyon Ralph3, (1)University of Idaho, Department of Computer Science, Moscow, ID, United States, (2)University of Idaho, Department of Computer Science, Moscow, United States, (3)mindat.org, Surray, United Kingdom
A Comprehensive Survey of Large‑Scale Language and Foundation Models Advancing Earth and Environmental Sciences (1988129)
Yitian Xiao1, Jieping Ye1, Grant Boquet1, James G Ogg2,3, Michael Stephenson4, Yu Ting1, Ting Jiang1, Mohamed Jaward Bah1, Yilin Li1, Hangyang Chen1 and Ying Zhou1, (1)Zhejiang Laboratory, Hangzhou City, China, (2)International Union of Geological Sciences, Deep-time Digital Earth, DDE Research Center of Excellence (Suzhou), Suzhou, China, (3)Purdue University, Geologic TimeScale Foundation, West Lafayette, United States, (4)Stephenson Geoscience Consulting Ltd, London, United Kingdom
A Geoscience Knowledge Graph Construction Method Based on Large Language Models (1990706)
Ting Yu1, Zhiqiang Feng2, Ting Jiang1, Wenxun Wu1, Mohamed Jaward Bah1, Hangyang Chen1, Jieping Ye1, Yilin Li1, Xianming Tang3 and Huiting Hu4, (1)Zhejiang Laboratory, Hangzhou City, China, (2)Zhejiang Center for Deep-time Digital Earth, Hangzhou City, China, (3)Petroleum Exploration and Production Research Institute, SINOPEC, Beijing, China, (4)Northeast Petroleum University, Daqing, China
