Dhruvit Patel, Troy Arcomano, Brian Hunt, Istvan Szunyogh, Edward Ott
{"title":"Prediction Beyond the Medium Range With an Atmosphere-Ocean Model That Combines Physics-Based Modeling and Machine Learning","authors":"Dhruvit Patel, Troy Arcomano, Brian Hunt, Istvan Szunyogh, Edward Ott","doi":"10.1029/2024MS004480","DOIUrl":"https://doi.org/10.1029/2024MS004480","url":null,"abstract":"<p>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.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"17 4","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024MS004480","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143836186","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rafael Santana, Guillaume Boutin, Christopher Horvat, Einar Ólason, Timothy Williams, Pierre Rampal
{"title":"Modeling Antarctic Sea Ice Variability Using a Brittle Rheology","authors":"Rafael Santana, Guillaume Boutin, Christopher Horvat, Einar Ólason, Timothy Williams, Pierre Rampal","doi":"10.1029/2024MS004584","DOIUrl":"https://doi.org/10.1029/2024MS004584","url":null,"abstract":"<p>Sea ice is a composite solid material that sustains large fracture features at scales from meters to kilometres. These fractures can play an important role in coupled atmosphere-ocean processes. To model these features, brittle sea ice physics, via the Brittle-Bingham-Maxwell (BBM) rheology, has been implemented in the Lagrangian neXt generation Sea Ice Model (neXtSIM). In Arctic-only simulations, the BBM rheology has shown a capacity to represent observationally consistent sea ice fracture patterns and breakup across a wide range of time and length scales. Still, it has not been tested whether this approach is suitable for the modeling of Antarctic sea ice, which is thinner and more seasonal compared to Arctic sea ice, and whether the ability to reproduce sea ice fractures has an impact on simulating Antarctic sea ice properties. Here, we introduce a new 50-km grid-spacing Antarctic configuration of neXtSIM, neXtSIM-Ant, using the BBM rheology. We evaluate this simulation against observations of sea ice extent, drift, and thickness and compare it with identically-forced neXtSIM simulations that use the standard modified Elastic-Visco-Plastic (mEVP) rheology. In general, using BBM results in thicker sea ice and an improved correlation of sea ice drift with observations than mEVP. We suggest that this is related to short-duration breakup events caused by Antarctic storms that are not well-simulated in the viscous-plastic model.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"17 4","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024MS004584","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143836074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Gunnar Behrens, Tom Beucler, Fernando Iglesias-Suarez, Sungduk Yu, Pierre Gentine, Michael Pritchard, Mierk Schwabe, Veronika Eyring
{"title":"Simulating Atmospheric Processes in Earth System Models and Quantifying Uncertainties With Deep Learning Multi-Member and Stochastic Parameterizations","authors":"Gunnar Behrens, Tom Beucler, Fernando Iglesias-Suarez, Sungduk Yu, Pierre Gentine, Michael Pritchard, Mierk Schwabe, Veronika Eyring","doi":"10.1029/2024MS004272","DOIUrl":"https://doi.org/10.1029/2024MS004272","url":null,"abstract":"<p>Deep learning is a powerful tool to represent subgrid processes in climate models, but many application cases have so far used idealized settings and deterministic approaches. Here, we develop stochastic parameterizations with calibrated uncertainty quantification to learn subgrid convective and turbulent processes and surface radiative fluxes of a superparameterization embedded in an Earth System Model (ESM). We explore three methods to construct stochastic parameterizations: (a) a single Deep Neural Network (DNN) with Monte Carlo Dropout; (b) a multi-member parameterization; and (c) a Variational Encoder Decoder with latent space perturbation. We show that the multi-member parameterization improves the representation of convective processes, especially in the planetary boundary layer, compared to individual DNNs. The respective uncertainty quantification illustrates that methods (b) and (c) are advantageous compared to a dropout-based DNN parameterization regarding the spread of convective processes. Hybrid simulations with our best-performing multi-member parameterizations remained challenging and crash within the first days. Therefore, we develop a pragmatic partial coupling strategy relying on the superparameterization for condensate emulation. Partial coupling reduces the computational efficiency of hybrid Earth-like simulations but enables model stability over 5 months with our multi-member parameterizations. However, our hybrid simulations exhibit biases in thermodynamic fields and differences in precipitation patterns. Despite this, the multi-member parameterizations enable improvements in reproducing tropical extreme precipitation compared to a traditional convection parameterization. Despite these challenges, our results indicate the potential of a new generation of multi-member machine learning parameterizations leveraging uncertainty quantification to improve the representation of stochasticity of subgrid effects.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"17 4","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024MS004272","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143826732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jerry Lin, Sungduk Yu, Liran Peng, Tom Beucler, Eliot Wong-Toi, Zeyuan Hu, Pierre Gentine, Margarita Geleta, Mike Pritchard
{"title":"Navigating the Noise: Bringing Clarity to ML Parameterization Design With \u0000 \u0000 \u0000 O\u0000 \u0000 $boldsymbol{mathcal{O}}$\u0000 (100) Ensembles","authors":"Jerry Lin, Sungduk Yu, Liran Peng, Tom Beucler, Eliot Wong-Toi, Zeyuan Hu, Pierre Gentine, Margarita Geleta, Mike Pritchard","doi":"10.1029/2024MS004551","DOIUrl":"https://doi.org/10.1029/2024MS004551","url":null,"abstract":"<p>Machine-learning (ML) parameterizations of subgrid processes (here of turbulence, convection, and radiation) may one day replace conventional parameterizations by emulating high-resolution physics without the cost of explicit simulation. However, uncertainty about the relationship between offline and online performance (i.e., when integrated with a large-scale general circulation model) hinders their development. Much of this uncertainty stems from limited sampling of the noisy, emergent effects of upstream ML design decisions on downstream online hybrid simulation. Our work rectifies the sampling issue via the construction of a semi-automated, end-to-end pipeline for <span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mi>O</mi>\u0000 <mrow>\u0000 <mo>(</mo>\u0000 <mn>100</mn>\u0000 <mo>)</mo>\u0000 </mrow>\u0000 </mrow>\u0000 <annotation> $mathcal{O}(100)$</annotation>\u0000 </semantics></math> size ensembles of hybrid simulations, revealing important nuances in how systematic reductions in offline error manifest in changes to online error and online stability. For example, removing dropout and switching from a Mean Squared Error to a Mean Absolute Error loss both reduce offline error, but they have opposite effects on online error and online stability. Other design decisions, like incorporating memory, converting moisture input from specific humidity to relative humidity, using batch normalization, and training on multiple climates do not come with any such compromises. Finally, we show that ensemble sizes of <span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mi>O</mi>\u0000 <mrow>\u0000 <mo>(</mo>\u0000 <mn>100</mn>\u0000 <mo>)</mo>\u0000 </mrow>\u0000 </mrow>\u0000 <annotation> $mathcal{O}(100)$</annotation>\u0000 </semantics></math> may be necessary to reliably detect causally relevant differences online. By enabling rapid online experimentation at scale, we can empirically settle debates regarding subgrid ML parameterization design that would have otherwise remained unresolved in the noise.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"17 4","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024MS004551","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143824633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}