Prediction Beyond the Medium Range With an Atmosphere‐Ocean Model That Combines Physics‐Based Modeling and Machine Learning

This paper explores the potential of a hybrid modeling approach that combines machine learning (ML) with conventional physics‐based modeling for weather prediction beyond the medium range. It extends the work of Arcomano et al. (2022, https://doi.org/10.1029/2021ms002712), which tested the approach...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	Journal of advances in modeling earth systems Ročník 17; číslo 4
Hlavní autori:	Patel, Dhruvit, Arcomano, Troy, Hunt, Brian, Szunyogh, Istvan, Ott, Edward
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	Washington John Wiley & Sons, Inc 01.04.2025
Predmet:	Atmospheric circulation Atmospheric circulation models Climate models Deep layer El Nino El Nino phenomena forecasting ENSO and teleconnections forecasting global precipitation Forecasting models General circulation models Heat content hybrid modeling Machine learning Madden-Julian oscillation Modelling Ocean models Oceans Parameterization Physics Planetary waves Precipitation Rossby waves Sea surface temperature subseasonal‐to‐seasonal Surface temperature Variables Weather forecasting weather forecasting beyond the medium range
ISSN:	1942-2466, 1942-2466
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Popis
Shrnutí:	This paper explores the potential of a hybrid modeling approach that combines machine learning (ML) with conventional physics‐based modeling for weather prediction beyond the medium range. It extends the work of Arcomano et al. (2022, https://doi.org/10.1029/2021ms002712), which tested the approach for short‐ and medium‐range weather prediction, and the work of Arcomano et al. (2023, https://doi.org/10.1029/2022gl102649), which investigated its potential for climate modeling. The hybrid model used for the forecast experiments of the paper is based on the low‐resolution, simplified parameterization atmospheric general circulation model SPEEDY. In addition to the hybridized prognostic variables of SPEEDY, the model has three purely ML‐based prognostic variables: the 6 hr cumulative precipitation, the sea surface temperature, and the heat content of the top 300 m deep layer of the ocean (a new addition compared to the model used in Arcomano et al., 2023, https://doi.org/10.1029/2022gl102649). The model has skill in predicting the El Niño cycle and its global teleconnections with precipitation for 3–7 months depending on the season. The model captures equatorial variability of the precipitation associated with Kelvin and Rossby waves and MJO. Predictions of the precipitation in the equatorial region have skill for 15 days in the East Pacific and 11.5 days in the West Pacific. Though the model has low spatial resolution, for these tasks it has prediction skill comparable to what has been published for high‐resolution, purely physics‐based, conventional, operational forecast models. Plain Language Summary We assess the potential of a hybrid modeling approach for weather prediction beyond 7–10 days. This approach combines machine learning (ML) with a conventional physics‐based model of the atmospheric general circulation. Some ML components of the hybrid model make corrections to the variables modeled by the conventional model. Other components provide information about the changes in the sea surface temperature that result from interactions between the atmosphere and ocean, or provide enhanced capabilities for the prediction of the precipitation. The results of forecast experiments with the model show that it has skill in predicting the El Niño cycle and its influence on the weather in other parts of the world for 3–7 months depending on the season. The model also provides useful forecast information about the precipitation in the Tropics for 11–15 days depending on the location. The tested approach could be used both to hybridize current high‐resolution purely physics‐based operational models and to enhance the capabilities of low‐resolution models of the general circulation used in academic research. Key Points A low‐resolution atmospheric general circulation model hybridized with machine learning has forecast skill beyond the medium range The low‐resolution hybrid model has prediction skill comparable to that of high‐resolution, purely physics‐based models for many tasks Compared to a conventional high‐resolution purely physics‐based approach, our hybrid approach requires much less computational resources
Bibliografia:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ISSN:	1942-2466 1942-2466
DOI:	10.1029/2024MS004480