利用 HEALPix 网格推进准深度学习天气预测

IF 4.4 2区 地球科学 Q1 METEOROLOGY & ATMOSPHERIC SCIENCES
Matthias Karlbauer, Nathaniel Cresswell-Clay, Dale R. Durran, Raul A. Moreno, Thorsten Kurth, Boris Bonev, Noah Brenowitz, Martin V. Butz
{"title":"利用 HEALPix 网格推进准深度学习天气预测","authors":"Matthias Karlbauer,&nbsp;Nathaniel Cresswell-Clay,&nbsp;Dale R. Durran,&nbsp;Raul A. Moreno,&nbsp;Thorsten Kurth,&nbsp;Boris Bonev,&nbsp;Noah Brenowitz,&nbsp;Martin V. Butz","doi":"10.1029/2023MS004021","DOIUrl":null,"url":null,"abstract":"<p>We present a parsimonious deep learning weather prediction model to forecast seven atmospheric variables with 3-hr time resolution for up to 1-year lead times on a 110-km global mesh using the Hierarchical Equal Area isoLatitude Pixelization (HEALPix). In comparison to state-of-the-art (SOTA) machine learning (ML) weather forecast models, such as Pangu-Weather and GraphCast, our DLWP-HPX model uses coarser resolution and far fewer prognostic variables. Yet, at 1-week lead times, its skill is only about 1 day behind both SOTA ML forecast models and the SOTA numerical weather prediction model from the European Center for Medium-Range Weather Forecasts. We report several improvements in model design, including switching from the cubed sphere to the HEALPix mesh, inverting the channel depth of the U-Net, and introducing gated recurrent units (GRU) on each level of the U-Net hierarchy. The consistent east-west orientation of all cells on the HEALPix mesh facilitates the development of location-invariant convolution kernels that successfully propagate weather patterns across the globe without requiring separate kernels for the polar and equatorial faces of the cube sphere. Without any loss of spectral power after the first 2 days, the model can be unrolled autoregressively for hundreds of steps into the future to generate realistic states of the atmosphere that respect seasonal trends, as showcased in 1-year simulations.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":null,"pages":null},"PeriodicalIF":4.4000,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004021","citationCount":"0","resultStr":"{\"title\":\"Advancing Parsimonious Deep Learning Weather Prediction Using the HEALPix Mesh\",\"authors\":\"Matthias Karlbauer,&nbsp;Nathaniel Cresswell-Clay,&nbsp;Dale R. Durran,&nbsp;Raul A. Moreno,&nbsp;Thorsten Kurth,&nbsp;Boris Bonev,&nbsp;Noah Brenowitz,&nbsp;Martin V. Butz\",\"doi\":\"10.1029/2023MS004021\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>We present a parsimonious deep learning weather prediction model to forecast seven atmospheric variables with 3-hr time resolution for up to 1-year lead times on a 110-km global mesh using the Hierarchical Equal Area isoLatitude Pixelization (HEALPix). In comparison to state-of-the-art (SOTA) machine learning (ML) weather forecast models, such as Pangu-Weather and GraphCast, our DLWP-HPX model uses coarser resolution and far fewer prognostic variables. Yet, at 1-week lead times, its skill is only about 1 day behind both SOTA ML forecast models and the SOTA numerical weather prediction model from the European Center for Medium-Range Weather Forecasts. We report several improvements in model design, including switching from the cubed sphere to the HEALPix mesh, inverting the channel depth of the U-Net, and introducing gated recurrent units (GRU) on each level of the U-Net hierarchy. The consistent east-west orientation of all cells on the HEALPix mesh facilitates the development of location-invariant convolution kernels that successfully propagate weather patterns across the globe without requiring separate kernels for the polar and equatorial faces of the cube sphere. Without any loss of spectral power after the first 2 days, the model can be unrolled autoregressively for hundreds of steps into the future to generate realistic states of the atmosphere that respect seasonal trends, as showcased in 1-year simulations.</p>\",\"PeriodicalId\":14881,\"journal\":{\"name\":\"Journal of Advances in Modeling Earth Systems\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":4.4000,\"publicationDate\":\"2024-08-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004021\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Advances in Modeling Earth Systems\",\"FirstCategoryId\":\"89\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1029/2023MS004021\",\"RegionNum\":2,\"RegionCategory\":\"地球科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"METEOROLOGY & ATMOSPHERIC SCIENCES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Advances in Modeling Earth Systems","FirstCategoryId":"89","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1029/2023MS004021","RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"METEOROLOGY & ATMOSPHERIC SCIENCES","Score":null,"Total":0}
引用次数: 0

摘要

我们提出了一种简洁的深度学习天气预报模型,利用层次等面积等纬度像素化(HEALPix)技术,在 110 千米的全球网格上以 3 小时的时间分辨率预报七个大气变量,预报周期最长可达 1 年。与最先进的(SOTA)机器学习(ML)天气预报模式(如盘古天气和 GraphCast)相比,我们的 DLWP-HPX 模式使用更粗糙的分辨率和更少的预报变量。然而,在 1 周的准备时间内,其技能仅比 SOTA ML 预报模式和欧洲中期天气预报中心的 SOTA 数值天气预报模式落后 1 天左右。我们报告了模型设计方面的几项改进,包括从立方球形网格转换到 HEALPix 网格,反转 U-Net 的通道深度,以及在 U-Net 层次结构的每一级引入门控循环单元(GRU)。HEALPix 网格上所有单元的东西方向一致,这有利于开发位置不变的卷积核,从而成功地将天气模式传播到全球各地,而无需为立方体球体的极地和赤道面分别建立核。在头两天之后,该模型的频谱功率不会有任何损失,可以自回归方式向未来展开数百步,生成尊重季节趋势的真实大气状态,这在 1 年模拟中得到了展示。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Advancing Parsimonious Deep Learning Weather Prediction Using the HEALPix Mesh

Advancing Parsimonious Deep Learning Weather Prediction Using the HEALPix Mesh

We present a parsimonious deep learning weather prediction model to forecast seven atmospheric variables with 3-hr time resolution for up to 1-year lead times on a 110-km global mesh using the Hierarchical Equal Area isoLatitude Pixelization (HEALPix). In comparison to state-of-the-art (SOTA) machine learning (ML) weather forecast models, such as Pangu-Weather and GraphCast, our DLWP-HPX model uses coarser resolution and far fewer prognostic variables. Yet, at 1-week lead times, its skill is only about 1 day behind both SOTA ML forecast models and the SOTA numerical weather prediction model from the European Center for Medium-Range Weather Forecasts. We report several improvements in model design, including switching from the cubed sphere to the HEALPix mesh, inverting the channel depth of the U-Net, and introducing gated recurrent units (GRU) on each level of the U-Net hierarchy. The consistent east-west orientation of all cells on the HEALPix mesh facilitates the development of location-invariant convolution kernels that successfully propagate weather patterns across the globe without requiring separate kernels for the polar and equatorial faces of the cube sphere. Without any loss of spectral power after the first 2 days, the model can be unrolled autoregressively for hundreds of steps into the future to generate realistic states of the atmosphere that respect seasonal trends, as showcased in 1-year simulations.

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
Journal of Advances in Modeling Earth Systems
Journal of Advances in Modeling Earth Systems METEOROLOGY & ATMOSPHERIC SCIENCES-
CiteScore
11.40
自引率
11.80%
发文量
241
审稿时长
>12 weeks
期刊介绍: The Journal of Advances in Modeling Earth Systems (JAMES) is committed to advancing the science of Earth systems modeling by offering high-quality scientific research through online availability and open access licensing. JAMES invites authors and readers from the international Earth systems modeling community. Open access. Articles are available free of charge for everyone with Internet access to view and download. Formal peer review. Supplemental material, such as code samples, images, and visualizations, is published at no additional charge. No additional charge for color figures. Modest page charges to cover production costs. Articles published in high-quality full text PDF, HTML, and XML. Internal and external reference linking, DOI registration, and forward linking via CrossRef.
×
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