[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-125672-en":3,"doc-seo-125672-105":30,"detail-sidebar-cat-0-en-105":95},{"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":20,"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},125672,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","Statistical Postprocessing of Numerical Weather Prediction Forecasts using Machine Learning - Dissertation","The dissertation develops statistical postprocessing methods for numerical weather prediction forecasts, focusing on improving probabilistic calibration and sharpness of ensemble outputs. It builds on theory for statistical forecasting, including forecast verification, proper scoring rules, consistent scoring functions, and representative forecast distribution families. The work proposes approaches for aggregating distribution forecasts from deep ensembles and evaluates established and neural-network-based postprocessing methods through simulation and case studies, including near real-time postprocessing and solar irradiance forecasting.","Statistical Postprocessing of Numerical Weather Prediction Forecasts using Machine Learning  \nZur Erlangung des akademischen Grades eines  \nDOKTORS DER NATURWISSENSCHAFTEN  \nvon der KIT-Fakultät für Mathematik des Karlsruher Instituts für Technologie (KIT) genehmigte  \nDISSERTATION  \nvon  \nBenedikt Schulz, M.Sc.  \ngeboren in Kandel  \nTag der mündlichen Prüfung: 17. Mai 2023  \nReferent: Prof. Dr. Tilmann Gneiting  \nKorreferenten: Prof. Dr. Peter Knippertz  \nDr. Sebastian Lerch  \nThis document is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0): [https://creativecommons.org/licenses/by/4.0/deed.en](https://creativecommons.org/licenses/by/4.0/deed.en)  \nAcknowledgments  \nMy deepest gratitude belongs to Sebastian Lerch, Peter Knippertz and Tilmann Gneiting for their excellent supervision over the last years. Through their guidance, support, experience and example, I was encouraged and able to pursue my doctoral studies, during which I have greatly enjoyed the interdisciplinary working environment being affiliated to research groups at three different faculties.  \nThe research presented in this thesis has been funded by the Deutsche Forschungsgemeinschaft (DFG) through the subproject “C5 – Dynamical feature-based ensemble postprocessing of wind gusts within European winter storms” of the Transregional Collaborative Research Center SFB / TRR 165 “Waves to Weather” and is gratefully acknowledged.  \nOver the course of my doctoral studies, I have benefited from the help and advice of many colleagues. First, I would like to thank Lea Eisenstein for being a great project partner in C5, whom I could always ask for advice and have fun with. Also, I would like to thank Eva-Maria Walz, Alexander Jordan, Johannes Resin, Ghulam Qadir, Daniel Wolffram, and all other members of the Computational Statistics group; Michael Maier-Gerber for an educational interdisciplinary collaboration, Marco Wurth for assistance on the KIT-Weather portal, Andreas Fink, and all other members of the working group Atmospheric Dynamics; Nina Horat, Jieyu Chen, and all other members of the Young Investigator Group “Artificial Intelligence for Probabilistic Weather Forecasting”; Tamara Göll for a semester full of thrilling teaching activities, Steffen Betsch, and all other members of the Institute for Stochastics; and Kevin Höhlein, and all early career scientists and members of Waves to Weather for many stimulating discussions and meetings.  \nFinally, I am grateful and feel very fortunate in light of the great support of my family and friends.  \nContents  \n1 Introduction 1  \n1.1 Relation to previous and published work . . . . . . . . . . . . . . . . . . . . . 3  \n2 Prelude: Theory on Statistical Forecasting 7  \n2.1 Prediction spaces, calibration and sharpness . . . . . . . . . . . . . . . . . . . 7  \n2.2 Forecast verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11  \n2.2.1 Proper scoring rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11  \n2.2.2 Consistent scoring functions ........................ 14  \n2.3 Exemplary types of forecast distributions .................... 15  \n2.3.1 Truncated and censored logistic distribution . . . . . . . . . . . . . . . 16  \n2.3.2 Bernstein quantile function   18  \n2.3.3 Piecewise uniform distribution . . . . . . . . . . . . . . . . . . . . . . 20  \n3 Aggregating Distribution Forecasts from Deep Ensembles 25  \n3.1 Combining predictive distributions ........................ 27  \n3.1.1 Linear pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27  \n3.1.2 Vincentization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28  \n3.2 Aggregating exemplary types of forecast distributions . . . . . . . . . . . . . 30  \n3.2. 1 Parametric forecast distribution . . . . . . . . . . . . . . . . . . . . . . 30  \n3.2.2 Bernstein quantile function   31  \n3.2.3 Piecewise uniform distribution . . . . . . . . . . . . . . . . . . . . . . 32  \n3.3 Simul","cbCaioBYIHb8YePW","https://ap.wps.com/l/cbCaioBYIHb8YePW","pdf",5125682,1,206,"English","en",105,"# Contents\n## Introduction\n## Prelude: Theory on Statistical Forecasting\n## Aggregating Distribution Forecasts from Deep Ensembles\n## Statistical Postprocessing: Methods\n## Statistical Postprocessing: Case Studies","[{\"question\":\"What problem does the dissertation address in numerical weather prediction?\",\"answer\":\"It addresses the need to statistically postprocess numerical weather prediction ensemble outputs to improve probabilistic forecast quality, such as calibration and sharpness.\"},{\"question\":\"Which theoretical foundations are covered before introducing the methods?\",\"answer\":\"The document reviews statistical forecasting theory, including prediction spaces, calibration and sharpness, forecast verification, proper scoring rules, consistent scoring functions, and examples of forecast distribution types.\"},{\"question\":\"How does the thesis combine forecasts from deep ensembles?\",\"answer\":\"It presents techniques for aggregating predictive distributions, including methods such as linear pooling and Vincentization, and discusses aggregation of representative forecast distribution types.\"},{\"question\":\"What empirical evaluations are included?\",\"answer\":\"It includes simulation and case studies, such as near real-time postprocessing on the KIT-Weather platform and solar irradiance forecasting, with details on data, model configurations, and results.\"}]","Statistical Postprocessing of Numerical Weather Prediction Forecasts using Machine Learning - Dissertation | PDF",1785900574,519,{"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":90,"head_meta":92,"extra_data":94,"updated_unix":28},"statistical-postprocessing-of-numerical-weather-prediction-forecasts-using-machine-learning-dissertation","",{"@graph":36,"@context":89},[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/statistical-postprocessing-of-numerical-weather-prediction-forecasts-using-machine-learning-dissertation/125672/",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":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81,85],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the dissertation address in numerical weather prediction?","Question",{"text":75,"@type":76},"It addresses the need to statistically postprocess numerical weather prediction ensemble outputs to improve probabilistic forecast quality, such as calibration and sharpness.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which theoretical foundations are covered before introducing the methods?",{"text":80,"@type":76},"The document reviews statistical forecasting theory, including prediction spaces, calibration and sharpness, forecast verification, proper scoring rules, consistent scoring functions, and examples of forecast distribution types.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the thesis combine forecasts from deep ensembles?",{"text":84,"@type":76},"It presents techniques for aggregating predictive distributions, including methods such as linear pooling and Vincentization, and discusses aggregation of representative forecast distribution types.",{"name":86,"@type":73,"acceptedAnswer":87},"What empirical evaluations are included?",{"text":88,"@type":76},"It includes simulation and case studies, such as near real-time postprocessing on the KIT-Weather platform and solar irradiance forecasting, with details on data, model configurations, and results.","https://schema.org",{"og:url":52,"og:type":91,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":93,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":96},[97,101,105,109,114,119,124,127,132,135,139],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":106,"show_sort_weight":107,"slug":108},"Exam",70,"exam",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},5,"Comic",60,"comic",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},6,"Technology",50,"technology",{"id":120,"doc_module":4,"doc_module_name":46,"category_name":121,"show_sort_weight":122,"slug":123},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":125,"slug":126},30,"research-report",{"id":128,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":130,"slug":131},9,"Religion & Spirituality",20,"religion-spirituality",{"id":130,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":130,"slug":134},"World Cup","world-cup",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":136,"slug":138},10,"Lifestyle","lifestyle",{"id":140,"doc_module":4,"doc_module_name":46,"category_name":141,"show_sort_weight":110,"slug":142},19,"General","general"]