[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126914-en":3,"doc-seo-126914-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},126914,2336474459895,"Aria","https://ap-avatar.wpscdn.com/avatar/22000baeef7a5ed0655?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786071322749376916",8,"Research & Report","Physics-Informed Machine Learning-Based Cloud Microphysics Parameterization for Earth System Models - ICON Model","Study develops a physics-informed machine learning (ML) cloud microphysics parameterization for the ICON Earth System Model. The ML scheme is trained on high-resolution simulation data to improve Earth System Model performance beyond traditional parameterizations. It explores a multilayer perceptron with feature engineering and physics constraints, supported by explainability methods to link inputs to outputs. Results show an R2 up to 0.777 for an individual feature and improved average R2 from 0.290 to 0.613 versus a baseline MLP.","PHYSICS-INFORMED MACHINE LEARNING-BASED CLOUD MICROPHYSICS PARAMETERIZATION FOR EARTH SYSTEM MODELS  \nEllen Sarauer  \nDeutsches Zentrum f¨ur Luft-und Raumfahrt Institut f¨ur Physik der Atmosphre Oberpfaffenhofen  \nGermany  \n[ellen.sarauer@dlr.de](ellen.sarauer@dlr.de)  \nMierk Schwabe  \nDeutsches Zentrum f¨ur Luft-und Raumfahrt Institut f¨ur Physik der Atmosphre Oberpfaffenhofen  \nGermany  \nPhilipp Weiss  \nAtmospheric, Oceanic and Planetary Physics  \nDepartment of Physics University of Oxford UK  \nAxel Lauer  \nDeutsches Zentrum f¨ur Luft-und Raumfahrt Institut f¨ur Physik der Atmosphre Oberpfaffenhofen  \nGermany  \nPhilip Stier  \nAtmospheric, Oceanic and Planetary Physics  \nDepartment of Physics University of Oxford UK  \nVeronika Eyring  \nDeutsches Zentrum f¨ur Luft-und Raumfahrt Institut f¨ur Physik der Atmosphre Oberpfaffenhofen  \nUniversity of Bremen  \nBremen  \nGermany  \nABSTRACT  \nIn this study, we develop a physics-informed machine learning (ML)-based cloud microphysics parameterization for the ICON model. By training the ML parameterization on high-resolution simulation data, we aim to improve Earth System Models (ESMs) in comparison to traditional parameterization schemes. We investigate the usage of a multilayer perceptron (MLP) with feature engineering and physics-constraints, and use explainability techniques to understand the relationship between input features and model output. Our novel approach yields promising results, with the physics-informed ML-based cloud microphysics parameterization achieving an R2 score up to 0 .777 for an individual feature. Additionally, we demonstrate a notable improvement in the overall performance in comparison to a baseline MLP, increasing its average R2 score from 0 .290 to 0 .613 across all variables. This approach to improve the representation of cloud microphysics in ESMs promises to enhance climate projections, contributing to a better understanding of climate change.  \n1 INTRODUCTION  \nMachine learning (ML) made great strides in enhancing Earth System Models (ESMs) by replacing traditional parameterizations that are based on empirical and physical understanding and represent the statistical effect of a given process at the grid scale of the climate model (Rasp et al., 2018; Reichstein et al., 2019; Eyring et al., 2023b) . For this, typically ML models are trained on high-resolution climate simulations or observations and then coupled to the coarse climate model, potentially eliminating long-standing biases in ESMs (Gentine et al., 2021; Grundner et al., 2022;  \nEyring et al., 2023a) . However, in the literature, the cloud microphysics parameterization has only been emulated on the same resolution as the simulation data (Gettelman et al., 2021; Han et al., 2020; Perkins et al., 2023) . Therefore, this study presents a novel approach for learning the cloud microphysics parameterization by incorporating higher resolution dynamics into the lower resolution ESM. This will be especially beneficial when parameterizing cloud convection together with cloud microphysics in future studies. The parameterization of cloud microphysics is central in traditional ESMs, working closely coupled to the convection scheme in order to model the behavior of clouds. It calculates the formation, growth and removal of cloud liquid water particles and captures phase changes. Additionally, cloud microphysics processes impact the lifetime of clouds, the water vapor distribution outside of clouds, the fluxes of water and radiation through the atmosphere, and latent heating. We aim to enhance the representation of subgrid-scale cloud microphysics within the coarse-scale Icosahedral Non-hydrostatic modeling framework (ICON) (Zngl et al., 2014; Giorgetta et al., 2018), which traditionally employs the single moment microphysical scheme (Lohmann & Roeckner, 1996) . This scheme focuses on prognostic equations for the mass mixing ratios (MMRs) of water vapor, cloud liquid water, cloud ice, and rain. In order to impr","cbCaivcavPA8qZXa","https://ap.wps.com/l/cbCaivcavPA8qZXa","pdf",11211229,1,11,"English","en",105,"# Abstract\n# Introduction\n# Methodology\n## Underlying Dataset","[{\"question\":\"What problem does the physics-informed ML approach address in Earth System Models?\",\"answer\":\"It targets biases and limitations of traditional cloud microphysics parameterizations by learning statistically consistent subgrid behavior from high-resolution simulation data and coupling it to the coarser ICON model.\"},{\"question\":\"How is the ML parameterization trained and constrained?\",\"answer\":\"It is trained on high-resolution ICON-based simulation data and uses a multilayer perceptron with feature engineering plus physics constraints to guide the learned relationship between inputs and cloud microphysical outputs.\"},{\"question\":\"What performance improvements does the new method achieve?\",\"answer\":\"The physics-informed ML approach yields an R2 up to 0.777 for an individual feature and increases the average R2 from 0.290 to 0.613 compared with a baseline MLP across variables.\"}]","Physics-Informed Machine Learning-Based Cloud Microphysics Parameterization for Earth System Models - ICON Model | PDF",1785935638,28,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"physics-informed-machine-learning-based-cloud-microphysics-parameterization-for-earth-system-models-icon-model","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/physics-informed-machine-learning-based-cloud-microphysics-parameterization-for-earth-system-models-icon-model/126914/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the physics-informed ML approach address in Earth System Models?","Question",{"text":75,"@type":76},"It targets biases and limitations of traditional cloud microphysics parameterizations by learning statistically consistent subgrid behavior from high-resolution simulation data and coupling it to the coarser ICON model.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How is the ML parameterization trained and constrained?",{"text":80,"@type":76},"It is trained on high-resolution ICON-based simulation data and uses a multilayer perceptron with feature engineering plus physics constraints to guide the learned relationship between inputs and cloud microphysical outputs.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance improvements does the new method achieve?",{"text":84,"@type":76},"The physics-informed ML approach yields an R2 up to 0.777 for an individual feature and increases the average R2 from 0.290 to 0.613 compared with a baseline MLP across variables.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]