Data Imbalance, Uncertainty Quantification, and Transfer Learning in Data-Driven Parameterizations: Lessons From the Emulation of Gravity Wave Momentum Transport in WACCM

dc.citation.articleNumbere2023MS004145
dc.citation.issueNumber7
dc.citation.journalTitleJournal of Advances in Modeling Earth Systems
dc.citation.volumeNumber16
dc.contributor.authorSun, Y. Qiang
dc.contributor.authorPahlavan, Hamid A.
dc.contributor.authorChattopadhyay, Ashesh
dc.contributor.authorHassanzadeh, Pedram
dc.contributor.authorLubis, Sandro W.
dc.contributor.authorAlexander, M. Joan
dc.contributor.authorGerber, Edwin P.
dc.contributor.authorSheshadri, Aditi
dc.contributor.authorGuan, Yifei
dc.date.accessioned2024-08-09T16:25:23Z
dc.date.available2024-08-09T16:25:23Z
dc.date.issued2024
dc.description.abstractNeural networks (NNs) are increasingly used for data-driven subgrid-scale parameterizations in weather and climate models. While NNs are powerful tools for learning complex non-linear relationships from data, there are several challenges in using them for parameterizations. Three of these challenges are (a) data imbalance related to learning rare, often large-amplitude, samples; (b) uncertainty quantification (UQ) of the predictions to provide an accuracy indicator; and (c) generalization to other climates, for example, those with different radiative forcings. Here, we examine the performance of methods for addressing these challenges using NN-based emulators of the Whole Atmosphere Community Climate Model (WACCM) physics-based gravity wave (GW) parameterizations as a test case. WACCM has complex, state-of-the-art parameterizations for orography-, convection-, and front-driven GWs. Convection- and orography-driven GWs have significant data imbalance due to the absence of convection or orography in most grid points. We address data imbalance using resampling and/or weighted loss functions, enabling the successful emulation of parameterizations for all three sources. We demonstrate that three UQ methods (Bayesian NNs, variational auto-encoders, and dropouts) provide ensemble spreads that correspond to accuracy during testing, offering criteria for identifying when an NN gives inaccurate predictions. Finally, we show that the accuracy of these NNs decreases for a warmer climate (4 × CO2). However, their performance is significantly improved by applying transfer learning, for example, re-training only one layer using ∼1% new data from the warmer climate. The findings of this study offer insights for developing reliable and generalizable data-driven parameterizations for various processes, including (but not limited to) GWs.
dc.identifier.citationSun, Y. Q., Pahlavan, H. A., Chattopadhyay, A., Hassanzadeh, P., Lubis, S. W., Alexander, M. J., Gerber, E. P., Sheshadri, A., & Guan, Y. (2024). Data Imbalance, Uncertainty Quantification, and Transfer Learning in Data-Driven Parameterizations: Lessons From the Emulation of Gravity Wave Momentum Transport in WACCM. Journal of Advances in Modeling Earth Systems, 16(7), e2023MS004145. https://doi.org/10.1029/2023MS004145
dc.identifier.digitalData-Imbalance
dc.identifier.doihttps://doi.org/10.1029/2023MS004145
dc.identifier.urihttps://hdl.handle.net/1911/117620
dc.language.isoeng
dc.publisherWiley
dc.rightsExcept where otherwise noted, this work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives (CC BY-NC-ND) license.  Permission to reuse, publish, or reproduce the work beyond the terms of the license or beyond the bounds of fair use or other exemptions to copyright law must be obtained from the copyright holder.
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.titleData Imbalance, Uncertainty Quantification, and Transfer Learning in Data-Driven Parameterizations: Lessons From the Emulation of Gravity Wave Momentum Transport in WACCM
dc.typeJournal article
dc.type.dcmiText
dc.type.publicationpublisher version
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Data-Imbalance.pdf
Size:
6.6 MB
Format:
Adobe Portable Document Format