date: 2026-09-03T11:52:56Z
pdf:PDFVersion: 1.3
pdf:docinfo:title: AI for atmosphere–ocean sciences: advancements, challenges and ways forward
xmp:CreatorTool: OUP
access_permission:can_print_degraded: true
subject: DOI: 10.1093/nsr/nwag063 National Science Review, 13, 0, 29-01-2026. Abstract: Artificial intelligence (AI) is rapidly transforming Earth science, offering unprecedented capabilities to tackle the most pressing challenges in the field. This work explores significant advances and emerging challenges across the AI for atmosphere–ocean sciences, while outlining critical ways forward. We review deep-learning methods and their application in weather and climate forecasting, which outperforms dynamical models in accuracy and computational efficiency. The role of AI in detecting complex phenomena, enhancing data assimilation and reconstruction, bias correction and downscaling coarse model outputs is also examined. However, the ‘black-box’ nature of complex AI models necessitates a focus on explainable AI to build trust and extract mechanistic insight. The most promising path forward is identified as the development of hybrid physics–AI modeling, which integrates the data-driven power of AI with the foundational constraints of physical laws to ensure generalizability and causal consistency. A new framework for AI-based model intercomparison is essential for rigorous benchmark performance. Finally, we contextualize these technical developments by discussing the usefulness and applicability of AI to society, including the improvement of multi-hazard early-warning systems and green energy production. We conclude by envisioning the future of AI agents for Earth science—autonomous, goal-oriented systems capable of designing and running experiments, generating and testing hypotheses, and learning dynamics from multisource data. This synthesis underscores that AI is not merely a tool, but a paradigm shift, which will significantly improve how we understand and adapt to a changing climate.
language: English
dc:format: application/pdf; version=1.3
pdf:docinfo:creator_tool: OUP
access_permission:fill_in_form: true
pdf:encrypted: false
dc:title: AI for atmosphere–ocean sciences: advancements, challenges and ways forward
modified: 2026-09-03T11:52:56Z
cp:subject: DOI: 10.1093/nsr/nwag063 National Science Review, 13, 0, 29-01-2026. Abstract: Artificial intelligence (AI) is rapidly transforming Earth science, offering unprecedented capabilities to tackle the most pressing challenges in the field. This work explores significant advances and emerging challenges across the AI for atmosphere–ocean sciences, while outlining critical ways forward. We review deep-learning methods and their application in weather and climate forecasting, which outperforms dynamical models in accuracy and computational efficiency. The role of AI in detecting complex phenomena, enhancing data assimilation and reconstruction, bias correction and downscaling coarse model outputs is also examined. However, the ‘black-box’ nature of complex AI models necessitates a focus on explainable AI to build trust and extract mechanistic insight. The most promising path forward is identified as the development of hybrid physics–AI modeling, which integrates the data-driven power of AI with the foundational constraints of physical laws to ensure generalizability and causal consistency. A new framework for AI-based model intercomparison is essential for rigorous benchmark performance. Finally, we contextualize these technical developments by discussing the usefulness and applicability of AI to society, including the improvement of multi-hazard early-warning systems and green energy production. We conclude by envisioning the future of AI agents for Earth science—autonomous, goal-oriented systems capable of designing and running experiments, generating and testing hypotheses, and learning dynamics from multisource data. This synthesis underscores that AI is not merely a tool, but a paradigm shift, which will significantly improve how we understand and adapt to a changing climate.
pdf:docinfo:subject: DOI: 10.1093/nsr/nwag063 National Science Review, 13, 0, 29-01-2026. Abstract: Artificial intelligence (AI) is rapidly transforming Earth science, offering unprecedented capabilities to tackle the most pressing challenges in the field. This work explores significant advances and emerging challenges across the AI for atmosphere–ocean sciences, while outlining critical ways forward. We review deep-learning methods and their application in weather and climate forecasting, which outperforms dynamical models in accuracy and computational efficiency. The role of AI in detecting complex phenomena, enhancing data assimilation and reconstruction, bias correction and downscaling coarse model outputs is also examined. However, the ‘black-box’ nature of complex AI models necessitates a focus on explainable AI to build trust and extract mechanistic insight. The most promising path forward is identified as the development of hybrid physics–AI modeling, which integrates the data-driven power of AI with the foundational constraints of physical laws to ensure generalizability and causal consistency. A new framework for AI-based model intercomparison is essential for rigorous benchmark performance. Finally, we contextualize these technical developments by discussing the usefulness and applicability of AI to society, including the improvement of multi-hazard early-warning systems and green energy production. We conclude by envisioning the future of AI agents for Earth science—autonomous, goal-oriented systems capable of designing and running experiments, generating and testing hypotheses, and learning dynamics from multisource data. This synthesis underscores that AI is not merely a tool, but a paradigm shift, which will significantly improve how we understand and adapt to a changing climate.
pdf:docinfo:creator: Luo Jing-Jia, Xia Jiangjiang, Pan Baoxiang, Ham Yoo-Geun, Li Xiaofeng, Shangguan Wei, Xue Wei, Wang Yaqiang, Mu Bin, Hong Youngjoon, Li Hao, Zhong Xiaohui, Dai Kan, Bai Lei, Ling Fenghua, Boers Niklas, Bretherton Christopher, Chen Bin, Cho Dongjin, Gentine Pierre, Guo Zijie, Huang Xiaomeng, Kang Daehyun, Kim Hyunwoo J., Kim Jeong-Hwan, Lei Lili, Meng Fan, Oh Seol-Hee, Qin Bo, Shen Zixiong, Sun Qiming, Tang Yuheng, Tong Xuan, Wan Bingcheng, Wang Lina, Wang Ya, Wang Yiming, Wu Jiye, Xiao Yi, Yao Lina, Yang Song, Yuan Chaoxia, Yuan Shijin, Yu Tingzhao, Zhao Mengchu
meta:author: Luo Jing-Jia, Xia Jiangjiang, Pan Baoxiang, Ham Yoo-Geun, Li Xiaofeng, Shangguan Wei, Xue Wei, Wang Yaqiang, Mu Bin, Hong Youngjoon, Li Hao, Zhong Xiaohui, Dai Kan, Bai Lei, Ling Fenghua, Boers Niklas, Bretherton Christopher, Chen Bin, Cho Dongjin, Gentine Pierre, Guo Zijie, Huang Xiaomeng, Kang Daehyun, Kim Hyunwoo J., Kim Jeong-Hwan, Lei Lili, Meng Fan, Oh Seol-Hee, Qin Bo, Shen Zixiong, Sun Qiming, Tang Yuheng, Tong Xuan, Wan Bingcheng, Wang Lina, Wang Ya, Wang Yiming, Wu Jiye, Xiao Yi, Yao Lina, Yang Song, Yuan Chaoxia, Yuan Shijin, Yu Tingzhao, Zhao Mengchu
meta:creation-date: 2026-03-10T07:39:17Z
created: 2026-03-10T07:39:17Z
access_permission:extract_for_accessibility: true
Creation-Date: 2026-03-10T07:39:17Z
pdf:docinfo:custom:doi: 10.1093/nsr/nwag063
Author: Luo Jing-Jia, Xia Jiangjiang, Pan Baoxiang, Ham Yoo-Geun, Li Xiaofeng, Shangguan Wei, Xue Wei, Wang Yaqiang, Mu Bin, Hong Youngjoon, Li Hao, Zhong Xiaohui, Dai Kan, Bai Lei, Ling Fenghua, Boers Niklas, Bretherton Christopher, Chen Bin, Cho Dongjin, Gentine Pierre, Guo Zijie, Huang Xiaomeng, Kang Daehyun, Kim Hyunwoo J., Kim Jeong-Hwan, Lei Lili, Meng Fan, Oh Seol-Hee, Qin Bo, Shen Zixiong, Sun Qiming, Tang Yuheng, Tong Xuan, Wan Bingcheng, Wang Lina, Wang Ya, Wang Yiming, Wu Jiye, Xiao Yi, Yao Lina, Yang Song, Yuan Chaoxia, Yuan Shijin, Yu Tingzhao, Zhao Mengchu
producer: Acrobat Distiller 10.0.0 (Windows); modified using iTextSharp.LGPLv2.Core 3.7.4.0
pdf:docinfo:producer: Acrobat Distiller 10.0.0 (Windows); modified using iTextSharp.LGPLv2.Core 3.7.4.0
doi: 10.1093/nsr/nwag063
pdf:unmappedUnicodeCharsPerPage: 0
dc:description: DOI: 10.1093/nsr/nwag063 National Science Review, 13, 0, 29-01-2026. Abstract: Artificial intelligence (AI) is rapidly transforming Earth science, offering unprecedented capabilities to tackle the most pressing challenges in the field. This work explores significant advances and emerging challenges across the AI for atmosphere–ocean sciences, while outlining critical ways forward. We review deep-learning methods and their application in weather and climate forecasting, which outperforms dynamical models in accuracy and computational efficiency. The role of AI in detecting complex phenomena, enhancing data assimilation and reconstruction, bias correction and downscaling coarse model outputs is also examined. However, the ‘black-box’ nature of complex AI models necessitates a focus on explainable AI to build trust and extract mechanistic insight. The most promising path forward is identified as the development of hybrid physics–AI modeling, which integrates the data-driven power of AI with the foundational constraints of physical laws to ensure generalizability and causal consistency. A new framework for AI-based model intercomparison is essential for rigorous benchmark performance. Finally, we contextualize these technical developments by discussing the usefulness and applicability of AI to society, including the improvement of multi-hazard early-warning systems and green energy production. We conclude by envisioning the future of AI agents for Earth science—autonomous, goal-oriented systems capable of designing and running experiments, generating and testing hypotheses, and learning dynamics from multisource data. This synthesis underscores that AI is not merely a tool, but a paradigm shift, which will significantly improve how we understand and adapt to a changing climate.
Keywords: AI application and challenge, atmosphere0ocean sciences, explainable AI, AI-MIP, AI agent
access_permission:modify_annotations: true
dc:creator: Luo Jing-Jia, Xia Jiangjiang, Pan Baoxiang, Ham Yoo-Geun, Li Xiaofeng, Shangguan Wei, Xue Wei, Wang Yaqiang, Mu Bin, Hong Youngjoon, Li Hao, Zhong Xiaohui, Dai Kan, Bai Lei, Ling Fenghua, Boers Niklas, Bretherton Christopher, Chen Bin, Cho Dongjin, Gentine Pierre, Guo Zijie, Huang Xiaomeng, Kang Daehyun, Kim Hyunwoo J., Kim Jeong-Hwan, Lei Lili, Meng Fan, Oh Seol-Hee, Qin Bo, Shen Zixiong, Sun Qiming, Tang Yuheng, Tong Xuan, Wan Bingcheng, Wang Lina, Wang Ya, Wang Yiming, Wu Jiye, Xiao Yi, Yao Lina, Yang Song, Yuan Chaoxia, Yuan Shijin, Yu Tingzhao, Zhao Mengchu
description: DOI: 10.1093/nsr/nwag063 National Science Review, 13, 0, 29-01-2026. Abstract: Artificial intelligence (AI) is rapidly transforming Earth science, offering unprecedented capabilities to tackle the most pressing challenges in the field. This work explores significant advances and emerging challenges across the AI for atmosphere–ocean sciences, while outlining critical ways forward. We review deep-learning methods and their application in weather and climate forecasting, which outperforms dynamical models in accuracy and computational efficiency. The role of AI in detecting complex phenomena, enhancing data assimilation and reconstruction, bias correction and downscaling coarse model outputs is also examined. However, the ‘black-box’ nature of complex AI models necessitates a focus on explainable AI to build trust and extract mechanistic insight. The most promising path forward is identified as the development of hybrid physics–AI modeling, which integrates the data-driven power of AI with the foundational constraints of physical laws to ensure generalizability and causal consistency. A new framework for AI-based model intercomparison is essential for rigorous benchmark performance. Finally, we contextualize these technical developments by discussing the usefulness and applicability of AI to society, including the improvement of multi-hazard early-warning systems and green energy production. We conclude by envisioning the future of AI agents for Earth science—autonomous, goal-oriented systems capable of designing and running experiments, generating and testing hypotheses, and learning dynamics from multisource data. This synthesis underscores that AI is not merely a tool, but a paradigm shift, which will significantly improve how we understand and adapt to a changing climate.
dcterms:created: 2026-03-10T07:39:17Z
Last-Modified: 2026-09-03T11:52:56Z
dcterms:modified: 2026-09-03T11:52:56Z
title: AI for atmosphere–ocean sciences: advancements, challenges and ways forward
xmpMM:DocumentID: uuid:b418ee9c-3bcb-36f2-aa21-a959afd99c6c
Last-Save-Date: 2026-09-03T11:52:56Z
pdf:docinfo:keywords: AI application and challenge, atmosphere0ocean sciences, explainable AI, AI-MIP, AI agent
pdf:docinfo:modified: 2026-09-03T11:52:56Z
meta:save-date: 2026-09-03T11:52:56Z
Content-Type: application/pdf
X-Parsed-By: org.apache.tika.parser.DefaultParser
creator: Luo Jing-Jia, Xia Jiangjiang, Pan Baoxiang, Ham Yoo-Geun, Li Xiaofeng, Shangguan Wei, Xue Wei, Wang Yaqiang, Mu Bin, Hong Youngjoon, Li Hao, Zhong Xiaohui, Dai Kan, Bai Lei, Ling Fenghua, Boers Niklas, Bretherton Christopher, Chen Bin, Cho Dongjin, Gentine Pierre, Guo Zijie, Huang Xiaomeng, Kang Daehyun, Kim Hyunwoo J., Kim Jeong-Hwan, Lei Lili, Meng Fan, Oh Seol-Hee, Qin Bo, Shen Zixiong, Sun Qiming, Tang Yuheng, Tong Xuan, Wan Bingcheng, Wang Lina, Wang Ya, Wang Yiming, Wu Jiye, Xiao Yi, Yao Lina, Yang Song, Yuan Chaoxia, Yuan Shijin, Yu Tingzhao, Zhao Mengchu
dc:language: English
dc:subject: AI application and challenge, atmosphere0ocean sciences, explainable AI, AI-MIP, AI agent
access_permission:assemble_document: true
xmpTPg:NPages: 27
pdf:charsPerPage: 4596
access_permission:extract_content: true
access_permission:can_print: true
meta:keyword: AI application and challenge, atmosphere0ocean sciences, explainable AI, AI-MIP, AI agent
access_permission:can_modify: true
pdf:docinfo:created: 2026-03-10T07:39:17Z