Correcting Systematic and State‐Dependent Errors in the NOAA FV3‐GFS Using Neural Networks

Correcting Systematic and State‐Dependent Errors in the NOAA FV3‐GFS Using Neural Networks
复制标题

使用神经网络纠正 NOAA FV3–GFS 中的系统性和状态相关错误

DOI:
10.1029/2022ms003309
复制
发表时间:
2022
影响因子:
6.8
通讯作者:
Tulich, Stefan
Tulich, Stefan
中科院分区:
地球科学2区
文献类型:
--
作者:
Chen, Tse‐Chun;Penny, Stephen G.;Whitaker, Jeffrey S.;Frolov, Sergey;Pincus, Robert;Tulich, Stefan

文献摘要

参考文献

被引文献

相似文献

用不完善的模型做出的天气预报包含与状态相关的误差。数据同化(DA)用观测的新信息部分地纠正了这些错误。同样,由数据处理过程产生的修正,或“分析增量”嵌入了关于模型错误的信息。本文试图提取这些信息以改进数值天气预报。神经网络(nn)被训练来预测基于大量分析增量的美国国家海洋和大气管理局FV3 - GFS模型中系统误差的修正。与线性基线相比,聚焦于大气柱的简单神经网络显著提高了估计模型的误差修正。与简单的列导向神经网络相比,使用卷积神经网络利用大规模水平流动条件并不能提高纠正模型误差的技能。模式误差修正对预报输入的敏感性受垂直水平和气象变量的高度局域化,且误差特征在垂直水平上存在差异。训练后,神经网络用于在模型集成期间对预测应用在线校正。改进将在循环数据分析系统和10天预测集合中进行评估。研究发现,将状态相关的神经网络预测修正应用于模型预测可以提高数据分析的整体质量,并在所有提前期提高10天的预测技能。
Weather forecasts made with imperfect models contain state‐dependent errors. Data assimilation (DA) partially corrects these errors with new information from observations. As such, the corrections, or “analysis increments,” produced by the DA process embed information about model errors. An attempt is made here to extract that information to improve numerical weather prediction. Neural networks (NNs) are trained to predict corrections to the systematic error in the National Oceanic and Atmospheric Administration's FV3‐GFS model based on a large set of analysis increments. A simple NN focusing on an atmospheric column significantly improves the estimated model error correction relative to a linear baseline. Leveraging large‐scale horizontal flow conditions using a convolutional NN, when compared to the simple column‐oriented NN, does not improve skill in correcting model error. The sensitivity of model error correction to forecast inputs is highly localized by vertical level and by meteorological variable, and the error characteristics vary across vertical levels. Once trained, the NNs are used to apply an online correction to the forecast during model integration. Improvements are evaluated both within a cycled DA system and across a collection of 10‐day forecasts. It is found that applying state‐dependent NN‐predicted corrections to the model forecast improves the overall quality of DA and improves the 10‐day forecast skill at all lead times.
UFS 天气模型
DOI: --
发表时间: 2020
期刊:
影响因子: --
作者:
Ufs Community
通讯作者: Ufs Community
深度学习估计可操作 NWP 同化系统中的模型偏差
DOI: --
发表时间: 2022
影响因子: 6.8
作者:
P. Laloyaux;T. Kurth;P. Dueben;D. Hall
通讯作者: D. Hall
DOI: 10.1017/eds.2022.7
发表时间: 2021-03
期刊: Environmental Data Science
影响因子: --
作者:
Antonios Mamalakis;I. Ebert‐Uphoff;E. Barnes
通讯作者: Antonios Mamalakis;I. Ebert‐Uphoff;E. Barnes
DOI: 10.1029/2020ms002232
发表时间: 2020-12-01
影响因子: 6.8
作者:
Bonavita, Massimo;Laloyaux, Patrick
通讯作者: Laloyaux, Patrick
DOI: 10.1029/2021ms002794
发表时间: 2021-09
影响因子: 6.8
作者:
C. Bretherton;B. Henn;Anna Kwa;Noah D. Brenowitz;Oliver Watt‐Meyer;J. McGibbon;W. A. Perkins;Spencer K. Clark;Lucas Harris
通讯作者: C. Bretherton;B. Henn;Anna Kwa;Noah D. Brenowitz;Oliver Watt‐Meyer;J. McGibbon;W. A. Perkins;Spencer K. Clark;Lucas Harris