利用神经网络建立 CO$_2$ 的大气传输模型

Vitus Benson, Ana Bastos, Christian Reimers, Alexander J. Winkler, Fanny Yang, Markus Reichstein
{"title":"利用神经网络建立 CO$_2$ 的大气传输模型","authors":"Vitus Benson, Ana Bastos, Christian Reimers, Alexander J. Winkler, Fanny Yang, Markus Reichstein","doi":"arxiv-2408.11032","DOIUrl":null,"url":null,"abstract":"Accurately describing the distribution of CO$_2$ in the atmosphere with\natmospheric tracer transport models is essential for greenhouse gas monitoring\nand verification support systems to aid implementation of international climate\nagreements. Large deep neural networks are poised to revolutionize weather\nprediction, which requires 3D modeling of the atmosphere. While similar in this\nregard, atmospheric transport modeling is subject to new challenges. Both,\nstable predictions for longer time horizons and mass conservation throughout\nneed to be achieved, while IO plays a larger role compared to computational\ncosts. In this study we explore four different deep neural networks (UNet,\nGraphCast, Spherical Fourier Neural Operator and SwinTransformer) which have\nproven as state-of-the-art in weather prediction to assess their usefulness for\natmospheric tracer transport modeling. For this, we assemble the CarbonBench\ndataset, a systematic benchmark tailored for machine learning emulators of\nEulerian atmospheric transport. Through architectural adjustments, we decouple\nthe performance of our emulators from the distribution shift caused by a steady\nrise in atmospheric CO$_2$. More specifically, we center CO$_2$ input fields to\nzero mean and then use an explicit flux scheme and a mass fixer to assure mass\nbalance. This design enables stable and mass conserving transport for over 6\nmonths with all four neural network architectures. In our study, the\nSwinTransformer displays particularly strong emulation skill (90-day $R^2 >\n0.99$), with physically plausible emulation even for forward runs of multiple\nyears. This work paves the way forward towards high resolution forward and\ninverse modeling of inert trace gases with neural networks.","PeriodicalId":501166,"journal":{"name":"arXiv - PHYS - Atmospheric and Oceanic Physics","volume":"20 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Atmospheric Transport Modeling of CO$_2$ with Neural Networks\",\"authors\":\"Vitus Benson, Ana Bastos, Christian Reimers, Alexander J. Winkler, Fanny Yang, Markus Reichstein\",\"doi\":\"arxiv-2408.11032\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Accurately describing the distribution of CO$_2$ in the atmosphere with\\natmospheric tracer transport models is essential for greenhouse gas monitoring\\nand verification support systems to aid implementation of international climate\\nagreements. Large deep neural networks are poised to revolutionize weather\\nprediction, which requires 3D modeling of the atmosphere. While similar in this\\nregard, atmospheric transport modeling is subject to new challenges. Both,\\nstable predictions for longer time horizons and mass conservation throughout\\nneed to be achieved, while IO plays a larger role compared to computational\\ncosts. In this study we explore four different deep neural networks (UNet,\\nGraphCast, Spherical Fourier Neural Operator and SwinTransformer) which have\\nproven as state-of-the-art in weather prediction to assess their usefulness for\\natmospheric tracer transport modeling. For this, we assemble the CarbonBench\\ndataset, a systematic benchmark tailored for machine learning emulators of\\nEulerian atmospheric transport. Through architectural adjustments, we decouple\\nthe performance of our emulators from the distribution shift caused by a steady\\nrise in atmospheric CO$_2$. More specifically, we center CO$_2$ input fields to\\nzero mean and then use an explicit flux scheme and a mass fixer to assure mass\\nbalance. This design enables stable and mass conserving transport for over 6\\nmonths with all four neural network architectures. In our study, the\\nSwinTransformer displays particularly strong emulation skill (90-day $R^2 >\\n0.99$), with physically plausible emulation even for forward runs of multiple\\nyears. This work paves the way forward towards high resolution forward and\\ninverse modeling of inert trace gases with neural networks.\",\"PeriodicalId\":501166,\"journal\":{\"name\":\"arXiv - PHYS - Atmospheric and Oceanic Physics\",\"volume\":\"20 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-08-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Atmospheric and Oceanic Physics\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2408.11032\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Atmospheric and Oceanic Physics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2408.11032","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

利用大气示踪传输模型准确描述大气中 CO$_2$ 的分布,对于温室气体监测和核查支持系统帮助实施国际气候协议至关重要。大型深度神经网络有望彻底改变天气预报,而天气预报需要对大气进行三维建模。大气传输建模与此类似,但也面临新的挑战。既要实现更长时间跨度的稳定预测,又要在整个过程中实现质量守恒,而与计算成本相比,IO 的作用更大。在这项研究中,我们探索了四种不同的深度神经网络(UNet、GraphCast、球形傅立叶神经运算器和 SwinTransformer),这些网络在天气预报中已被证明是最先进的,以评估它们在大气示踪剂传输建模中的实用性。为此,我们建立了 CarbonBenchdataset 数据集,这是一个为欧勒大气传输机器学习模拟器量身定制的系统基准。通过结构调整,我们将仿真器的性能与大气中 CO$_2$ 稳步上升引起的分布偏移分离开来。更具体地说,我们将 CO$_2$ 输入场的中心设定为零均值,然后使用显式通量方案和质量固定器来确保质量平衡。这种设计使所有四种神经网络架构都能在 6 个月内实现稳定的质量保证传输。在我们的研究中,SwinTransformer 显示出了特别强的模拟能力(90 天的 R^2 >0.99),甚至在多年的前向运行中也具有物理上可信的模拟能力。这项工作为利用神经网络对惰性痕量气体进行高分辨率正演和反演建模铺平了道路。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Atmospheric Transport Modeling of CO$_2$ with Neural Networks
Accurately describing the distribution of CO$_2$ in the atmosphere with atmospheric tracer transport models is essential for greenhouse gas monitoring and verification support systems to aid implementation of international climate agreements. Large deep neural networks are poised to revolutionize weather prediction, which requires 3D modeling of the atmosphere. While similar in this regard, atmospheric transport modeling is subject to new challenges. Both, stable predictions for longer time horizons and mass conservation throughout need to be achieved, while IO plays a larger role compared to computational costs. In this study we explore four different deep neural networks (UNet, GraphCast, Spherical Fourier Neural Operator and SwinTransformer) which have proven as state-of-the-art in weather prediction to assess their usefulness for atmospheric tracer transport modeling. For this, we assemble the CarbonBench dataset, a systematic benchmark tailored for machine learning emulators of Eulerian atmospheric transport. Through architectural adjustments, we decouple the performance of our emulators from the distribution shift caused by a steady rise in atmospheric CO$_2$. More specifically, we center CO$_2$ input fields to zero mean and then use an explicit flux scheme and a mass fixer to assure mass balance. This design enables stable and mass conserving transport for over 6 months with all four neural network architectures. In our study, the SwinTransformer displays particularly strong emulation skill (90-day $R^2 > 0.99$), with physically plausible emulation even for forward runs of multiple years. This work paves the way forward towards high resolution forward and inverse modeling of inert trace gases with neural networks.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信