{"title":"Calibrating Tropical Forest Coexistence in Ecosystem Demography Models Using Multi-Objective Optimization Through Population-Based Parallel Surrogate Search","authors":"Yanyan Cheng, Wenyu Wang, Matteo Detto, Rosie Fisher, Christine Shoemaker","doi":"10.1029/2023MS004195","DOIUrl":"https://doi.org/10.1029/2023MS004195","url":null,"abstract":"<p>Tropical forest diversity governs forest structures, compositions, and influences the ecosystem response to environmental changes. Better representation of forest diversity in ecosystem demography (ED) models within Earth system models is thus necessary to accurately capture and predict how tropical forests affect Earth system dynamics subject to climate changes. However, achieving forest coexistence in ED models is challenging due to their computational expense and limited understanding of the mechanisms governing forest functional diversity. This study applies the advanced Multi-Objective Population-based Parallel Local Surrogate-assisted search (MOPLS) optimization algorithm to simultaneously calibrate ecosystem fluxes and coexistence of two physiologically distinct tropical forest species in a size- and age-structured ED model with realistic representation of wood harvest. MOPLS exhibits satisfactory model performance, capturing hydrological and biogeochemical dynamics observed in Barro Colorado Island, Panama, and robustly achieving coexistence for the two representative forest species. This demonstrates its effectiveness in calibrating tropical forest coexistence. The optimal solution is applied to investigate the recovery trajectories of forest biomass after various intensities of clear-cut deforestation. We find that a 20% selective logging can take approximately 40 years for aboveground biomass to return to the initial level. This is due to the slow recovery rate of late successional trees, which only increases by 4% over the 40-year period. This study lays the foundation to calibrate coexistence in ED models. MOPLS can be an effective tool to help better represent tropical forest diversity in Earth system models and inform forest management practices.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004195","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141967688","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}
{"title":"On Energy-Aware Hybrid Models","authors":"Igor Shevchenko, Dan Crisan","doi":"10.1029/2024MS004306","DOIUrl":"https://doi.org/10.1029/2024MS004306","url":null,"abstract":"<p>This study proposes deterministic and stochastic energy-aware hybrid models that should enable simulations of idealized and primitive-equations Geophysical Fluid Dynamics (GFD) models at low resolutions without compromising on quality compared with high-resolution runs. Such hybrid models bridge the data-driven and physics-driven modeling paradigms by combining regional stability and classical GFD models at low resolution that cannot reproduce high-resolution reference flow features (large-scale flows and small-scale vortices) which are, however, resolved. Hybrid models use an energy-aware correction of advection velocity and extra forcing compensating for the drift of the low-resolution model away from the reference phase space. The main advantages of hybrid models are that they allow for physics-driven flow recombination within the reference energy band, reproduce resolved reference flow features, and produce more accurate ensemble forecasts than their classical GFD counterparts. Hybrid models offer appealing benefits and flexibility to the modeling and forecasting communities, as they are computationally cheap and can use both numerically-computed flows and observations from different sources. All these suggest that the hybrid approach has the potential to exploit low-resolution models for long-term weather forecasts and climate projections thus offering a new cost effective way of GFD modeling. The proposed hybrid approach has been tested on a three-layer quasi-geostrophic model for a beta-plane Gulf Stream flow configuration. The results show that the low-resolution hybrid model reproduces the reference flow features that are resolved on the coarse grid and also gives a more accurate ensemble forecast than the physics-driven model.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024MS004306","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141967687","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}
Mohamad Abed El Rahman Hammoud, Naila Raboudi, Edriss S. Titi, Omar Knio, Ibrahim Hoteit
{"title":"Data Assimilation in Chaotic Systems Using Deep Reinforcement Learning","authors":"Mohamad Abed El Rahman Hammoud, Naila Raboudi, Edriss S. Titi, Omar Knio, Ibrahim Hoteit","doi":"10.1029/2023MS004178","DOIUrl":"https://doi.org/10.1029/2023MS004178","url":null,"abstract":"<p>Data assimilation (DA) plays a pivotal role in diverse applications, ranging from weather forecasting to trajectory planning for autonomous vehicles. A prime example is the widely used ensemble Kalman filter (EnKF), which relies on the Kalman filter's linear update equation to correct each of the ensemble forecast member's state with incoming observations. Recent advancements have witnessed the emergence of deep learning approaches in this domain, primarily within a supervised learning framework. However, the adaptability of such models to untrained scenarios remains a challenge. In this study, we introduce a new DA strategy that utilizes reinforcement learning (RL) to apply state corrections using full or partial observations of the state variables. Our investigation focuses on demonstrating this approach to the chaotic Lorenz 63 and 96 systems, where the agent's objective is to maximize the geometric series with terms that are proportional to the negative root-mean-squared error (RMSE) between the observations and corresponding forecast states. Consequently, the agent develops a correction strategy, enhancing model forecasts based on available observations. Our strategy employs a stochastic action policy, enabling a Monte Carlo-based DA framework that relies on randomly sampling the policy to generate an ensemble of assimilated realizations. Numerical results demonstrate that the developed RL algorithm performs favorably when compared to the EnKF. Additionally, we illustrate the agent's capability to assimilate non-Gaussian observations, addressing one of the limitations of the EnKF.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004178","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141967673","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}
{"title":"A Machine Learning Bias Correction on Large-Scale Environment of High-Impact Weather Systems in E3SM Atmosphere Model","authors":"Shixuan Zhang, Bryce Harrop, L. Ruby Leung, Alexis-Tzianni Charalampopoulos, Benedikt Barthel Sorensen, Wenwei Xu, Themistoklis Sapsis","doi":"10.1029/2023MS004138","DOIUrl":"https://doi.org/10.1029/2023MS004138","url":null,"abstract":"<p>Large-scale dynamical and thermodynamical processes are common environmental drivers of high-impact weather systems causing extreme weather events. However, such large-scale environmental conditions often display systematic biases in climate simulations, posing challenges to evaluating high-impact weather systems and extreme weather events. In this paper, a machine learning (ML) approach was employed to bias correct the large-scale wind, temperature, and humidity simulated by the atmospheric component of the Energy Exascale Earth System Model (E3SM) at ∼1° resolution. The usefulness of the ML approach for extreme weather analysis was demonstrated with a focus on three high-impact weather systems, including tropical cyclones (TCs), extratropical cyclones (ETCs), and atmospheric rivers (ARs). We show that the ML model can effectively reduce climate bias in large-scale wind, temperature, and humidity while preserving their responses to imposed climate change perturbations. The bias correction is found to directly improve water vapor transport associated with ARs, and representations of thermodynamical flows associated with ETCs. When the bias-corrected large-scale winds are used to drive a synthetic TC track forecast model over the Atlantic basin, the resulting TC track density agrees better with that of the TC track model driven by observed winds. In addition, the ML model insignificantly interferes with the mean climate change signals of large-scale storm environments as well as the occurrence and intensity of three weather systems. This study suggests that the proposed ML approach can be used to improve the downscaling of extreme weather events by providing more realistic large-scale storm environments simulated by low-resolution climate models.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004138","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141967689","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}
Gareth S. Jones, Martin B. Andrews, Timothy Andrews, Ed Blockley, Andrew Ciavarella, Nikos Christidis, Daniel F. Cotterill, Fraser C. Lott, Jeff Ridley, Peter A. Stott
{"title":"The HadGEM3-GC3.1 Contribution to the CMIP6 Detection and Attribution Model Intercomparison Project","authors":"Gareth S. Jones, Martin B. Andrews, Timothy Andrews, Ed Blockley, Andrew Ciavarella, Nikos Christidis, Daniel F. Cotterill, Fraser C. Lott, Jeff Ridley, Peter A. Stott","doi":"10.1029/2023MS004135","DOIUrl":"https://doi.org/10.1029/2023MS004135","url":null,"abstract":"<p>The UK contribution to the Detection and Attribution Model Intercomparison Project (DAMIP), part of the sixth phase of the Climate Model Intercomparison Project (CMIP6), is described. The lower atmosphere and ocean resolution configuration of the latest Hadley Centre global environmental model, HadGEM3-GC3.1, is used to create simulations driven either with historical changes in anthropogenic well-mixed greenhouse gases, anthropogenic aerosols, or natural climate factors. Global mean near-surface air temperatures from the HadGEM3-GC31-LL simulations are consistent with CMIP6 model ensembles for the equivalent experiments. While the HadGEM3-GC31-LL simulations with anthropogenic and natural forcing factors capture the overall observed warming, the lack of marked simulated warming until the 1990s is diagnosed as due to aerosol cooling mostly offsetting the well-mixed greenhouse gas warming until then. The model has unusual temperature variability over the Southern Ocean related to occasional deep convection bringing heat to the surface. This is most prominent in the model's aerosol only simulations, which have the curious feature of warming in the high southern latitudes, while the rest of the globe cools, a behavior not seen in other CMIP6 models. This has implications for studies that assume model responses, from different climate drivers, can be linearly combined. While DAMIP was predominantly designed for detection and attribution studies, the experiments are also very valuable for understanding how different climate drivers influence a model, and thus for interpretating the responses of combined anthropogenic and natural driven simulations. We recommend institutions provide model simulations for the high priority DAMIP experiments.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004135","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141966566","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}
{"title":"An Eigenvalue-Based Framework for Constraining Anisotropic Eddy Viscosity","authors":"Scott D. Bachman","doi":"10.1029/2024MS004375","DOIUrl":"https://doi.org/10.1029/2024MS004375","url":null,"abstract":"<p>Eddy viscosity is employed throughout the majority of numerical fluid dynamical models, and has been the subject of a vigorous body of research spanning a variety of disciplines. It has long been recognized that the proper description of eddy viscosity uses tensor mathematics, but in practice it is almost always employed as a scalar due to uncertainty about how to constrain the extra degrees of freedom and physical properties of its tensorial form. This manuscript borrows techniques from outside the realm of geophysical fluid dynamics to consider the eddy viscosity tensor using its eigenvalues and eigenvectors, establishing a new framework by which tensorial eddy viscosity can be tested. This is made possible by a careful analysis of an operation called tensor unrolling, which casts the eigenvalue problem for a fourth-order tensor into a more familiar matrix-vector form, whereby it becomes far easier to understand and manipulate. New constraints are established for the eddy viscosity coefficients that are guaranteed to result in energy dissipation, backscatter, or a combination of both. Finally, a testing protocol is developed by which tensorial eddy viscosity can be systematically evaluated across a wide range of fluid regimes.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024MS004375","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141966641","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}
Yunpeng Shan, Jiwen Fan, Kai Zhang, Jacob Shpund, Christopher Terai, Guang J. Zhang, Xiaoliang Song, Chih-Chieh-Jack Chen, Wuyin Lin, Xiaohong Liu, Manish Shrivastava, Hailong Wang, Shaocheng Xie
{"title":"Improving Aerosol Radiative Forcing and Climate in E3SM: Impacts of New Cloud Microphysics and Improved Wet Removal Treatments","authors":"Yunpeng Shan, Jiwen Fan, Kai Zhang, Jacob Shpund, Christopher Terai, Guang J. Zhang, Xiaoliang Song, Chih-Chieh-Jack Chen, Wuyin Lin, Xiaohong Liu, Manish Shrivastava, Hailong Wang, Shaocheng Xie","doi":"10.1029/2023MS004059","DOIUrl":"https://doi.org/10.1029/2023MS004059","url":null,"abstract":"<p>Numerous Earth system models exhibit excessive aerosol effective forcing at the top of the atmosphere (TOA), including the Department of Energy's Energy Exascale Earth System Model (E3SM). Here, in the context of the E3SM version 3 effort, the predicted particle property (P3) stratiform cloud microphysics scheme and an enhanced deep convection parameterization suite (ZM_plus) are implemented into E3SM. The ZM_plus includes a convective cloud microphysics scheme, a multi-scale coherent structure parameterization for mesoscale convective systems, and a revised cloud base mass flux formulation considering impacts of the large-scale environment. The P3 scheme improved cloud and radiation particularly over the Northern Hemisphere and the frequency of heavy precipitation over the tropics, and the ZM_plus improved clouds in the tropics. P3 decreases aerosol effective forcing by 0.15 W m<sup>−2</sup>, while the ZM_plus increases it by 0.27 W m<sup>−2</sup>, resulting from excessive direct (0.31 W m<sup>−2</sup>) and indirect forcing (−1.79 W m<sup>−2</sup>). The excessive aerosol forcings are due to aerosol overestimation associated with insufficient aerosol wet removal. By improving the physical treatments in the aerosol wet removal, we effectively mitigate anthropogenic aerosol overestimation and thus attenuate direct (0.09 W m<sup>−2</sup>) and indirect aerosol forcing (−1.52 W m<sup>−2</sup>). Adjustment to primary organic matter hygroscopicity reduces direct and indirect forcing to more reasonable values: −0.13 W m<sup>−2</sup> and −1.31 W m<sup>−2</sup>, respectively. On climatology, improved aerosol treatments mitigate overestimation of aerosol optical depth.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004059","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141968437","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}
{"title":"A Comparison of Diagnostics for AMOC Heat Transport Applied to the CESM Large Ensemble","authors":"C Spencer Jones, Scout Jiang, Ryan P. Abernathey","doi":"10.1029/2023MS003978","DOIUrl":"https://doi.org/10.1029/2023MS003978","url":null,"abstract":"<p>Atlantic time-mean heat transport is northward at all latitudes and exhibits strong multidecadal variability between about 30°N and 55°N. Atlantic heat transport variability influences many aspects of the climate system, including regional surface temperatures, subpolar heat content, Arctic sea-ice concentration and tropical precipitation patterns. Atlantic heat transport and heat transport variability are commonly partitioned into two components: the heat transport by the Atlantic Meridional Overturning Circulation (AMOC) and the heat transport by the gyres. In this paper we compare four different methods for performing this partition, and we apply these methods to the Community Earth System Model Large Ensemble at 34°N, 26°N and 5°S. We discuss the strengths and weaknesses of each method. The four methods all give significantly different estimates for the proportion of the time-mean heat transport performed by AMOC. One of these methods is a new physically-motivated method based on the pathway of the northward-flowing part of AMOC. This paper presents a preliminary version of our method that works only when the AMOC follows the western boundary of the basin. All the methods agree that at 26°N, 80%–100% of heat transport variability at 2–10 years timescales is performed by AMOC, but there is more disagreement between methods in attributing multidecadal variability, with some methods showing a compensation between the AMOC and gyre heat transport variability.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 8","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS003978","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141968340","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}
Sajjad Azimi, Anna Jaruga, Emily de Jong, Sylwester Arabas, Tapio Schneider
{"title":"Training Warm-Rain Bulk Microphysics Schemes Using Super-Droplet Simulations","authors":"Sajjad Azimi, Anna Jaruga, Emily de Jong, Sylwester Arabas, Tapio Schneider","doi":"10.1029/2023MS004028","DOIUrl":"10.1029/2023MS004028","url":null,"abstract":"<p>Cloud microphysics is a critical aspect of the Earth's climate system, which involves processes at the nano- and micrometer scales of droplets and ice particles. In climate modeling, cloud microphysics is commonly represented by bulk models, which contain simplified process rates that require calibration. This study presents a framework for calibrating warm-rain bulk schemes using high-fidelity super-droplet simulations that provide a more accurate and physically based representation of cloud and precipitation processes. The calibration framework employs ensemble Kalman methods including Ensemble Kalman Inversion and Unscented Kalman Inversion to calibrate bulk microphysics schemes with probabilistic super-droplet simulations. We demonstrate the framework's effectiveness by calibrating a single-moment bulk scheme, resulting in a reduction of data-model mismatch by more than 75% compared to the model with initial parameters. Thus, this study demonstrates a powerful tool for enhancing the accuracy of bulk microphysics schemes in atmospheric models and improving climate modeling.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 7","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004028","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141847706","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}
Y. Qiang Sun, Hamid A. Pahlavan, Ashesh Chattopadhyay, Pedram Hassanzadeh, Sandro W. Lubis, M. Joan Alexander, Edwin P. Gerber, Aditi Sheshadri, Yifei Guan
{"title":"Data Imbalance, Uncertainty Quantification, and Transfer Learning in Data-Driven Parameterizations: Lessons From the Emulation of Gravity Wave Momentum Transport in WACCM","authors":"Y. Qiang Sun, Hamid A. Pahlavan, Ashesh Chattopadhyay, Pedram Hassanzadeh, Sandro W. Lubis, M. Joan Alexander, Edwin P. Gerber, Aditi Sheshadri, Yifei Guan","doi":"10.1029/2023MS004145","DOIUrl":"10.1029/2023MS004145","url":null,"abstract":"<p>Neural 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 × CO<sub>2</sub>). 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.</p>","PeriodicalId":14881,"journal":{"name":"Journal of Advances in Modeling Earth Systems","volume":"16 7","pages":""},"PeriodicalIF":4.4,"publicationDate":"2024-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2023MS004145","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141847359","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}