[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-118566-en":3,"doc-seo-118566-105":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},118566,13056703019662,"Evangeline","https://ap-avatar.wpscdn.com/avatar/be000253a8e92610077?_k=1778726343310543188",8,"Research & Report","Fast Flood Modelling after Dike Breaches: Machine Learning or Conceptual Models?","Two-dimensional depth-averaged hydrodynamic models support flood consequence analysis, but their long computation times hinder real-time forecasting. This study builds and compares a machine learning surrogate and a conceptual surrogate to predict overland flow propagation following a dike breach. The machine learning model achieves about 95% accuracy, while the conceptual model reaches roughly 70–80%. Trade-offs for emergency-time deployment are evaluated, showing both approaches can increasingly support time-sensitive flood decision-making.","Fast flood modelling after dike breaches: machine learning or conceptual models?  \nL.S. Besseling, A. Bomers & S.J.M.H. Hulscher University of Twente, Enschede, The Netherlands  \nABSTRACT: Two-dimensional depth-averaged (2DH) hydrodynamic models are often used to obtain insight in the consequences of a flood. However, the long computation times of these models do not allow for real-time flood forecasting. Surrogate models such as machine learning models or conceptual models promise much lower computation times while achieving reasonable accuracy. In this study, we develop and compare a machine learning model and a conceptual model for the prediction of overland flow propagation after a dike breach. We find a high accuracy of around 95% by the machine learning model, and a reasonable accuracy of around 70– 80% by the conceptual model. We discuss trade-offs of each method regarding practical use ina real-time flood forecasting system, and conclude that both machine learning and conceptual models have a growing capability to be applied during time-sensitive emergency situations.  \n1 INTRODUCTION  \nDike breach floods are a risk along many of the world’s major rivers, and they can have sudden and severe consequences. Due to the unexpected nature and often high densities of people and economic activity in the hinterlands, prediction of these events and their consequences is vital.  \nFlood models are important to obtain insights in these types of floods. Often, hydrodynamic models with a one-dimensional (1D) part for the river channel and a two-dimensional depthaveraged (2DH) part for the hinterland are used. However, these models are too computationally expensive for scenario analysis during an emergency situation. For real-time flood forecasting, ensemble model predictions and uncertainty analysis enable authorities to make optimal decisions (Chu et al. 2020). Therefore, faster modelling techniques are desired.  \nSurrogate models are cheaper-to-run models that aim to approximate the original model’s behaviour (Razavi et al. 2012) . For example, response surface surrogates are models that use data-driven approximations of the original data or model. Neural networks are the most common type of response surface surrogate, and function by learning input-output relations in a dataset (Mosavi et al. 2018) . Another type of surrogate models is conceptual models, which do not include physical processes but rather rely on simplified hydraulic concepts (Teng et al. 2017) . For example, flood models such as the Height Above Nearest Drainage (HAND) model use only the elevation map to determine flood extent and depths (Nobre et al. 2011) .  \nBoth neural networks and the HAND model have been researched in the context of river floods, but is it unknown how these models perform for dike breach flood modelling. Dike breaches result in different flow behaviour near to the breach (fast rising and high water depths) compared to far from the breach (slower rising and lower water depths) . This study develops aversion of both these model types suitable for dike breach flood modelling, and assesses their accuracy and practicality for use in a real-time flood forecasting system.  \n2 METHODOLOGY  \nWe develop a neural network and a HAND-based conceptual model to predict the overland flow propagation after a dike breach. A dataset of a 1D2D hydrodynamic model constructed in HECRAS by Bomers et al. (2019) is available for this research. For information on the model, we refer to Bomers et al. (2019) . The model contains the Dutch part of the Rhine river, from the Dutch-German border up to and including the main distributaries: the Waal, Nederrijn and IJssel (Fig. 1) . Bomers (2021) use this model to train a neural network to predict the outflow hydrograph of dike breaches in this bifurcating system. Data of 73 different extreme discharge waves from this study are available, and at one location the dike breached during all these simulations. This breach locatio","cbCaif720ECaX8lq","https://ap.wps.com/l/cbCaif720ECaX8lq","pdf",14629370,1,7,"English","en",105,"# Abstract\n# Introduction\n## Surrogate models and conceptual approaches\n# Methodology\n## Neural network\n## Conceptual model (HAND)\n## Dataset and study setup","[{\"question\":\"Why are faster models needed for dike breach flood forecasting?\",\"answer\":\"Hydrodynamic 2DH models are computationally expensive, making them unsuitable for scenario analysis and real-time forecasting during emergencies.\"},{\"question\":\"What surrogate models are compared in the study?\",\"answer\":\"A machine learning model based on a neural network (LSTM) and a conceptual model based on HAND-style simplified hydraulic concepts are developed and compared.\"},{\"question\":\"How accurate are the two approaches for predicting overland flow propagation?\",\"answer\":\"The machine learning model reaches around 95% accuracy, while the conceptual model achieves about 70–80% accuracy.\"}]","Fast Flood Modelling after Dike Breaches: Machine Learning or Conceptual Models? | PDF",1785684264,18,{"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},"fast-flood-modelling-after-dike-breaches-machine-learning-or-conceptual-models","",{"@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/fast-flood-modelling-after-dike-breaches-machine-learning-or-conceptual-models/118566/",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","2026-08-02",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},"Why are faster models needed for dike breach flood forecasting?","Question",{"text":75,"@type":76},"Hydrodynamic 2DH models are computationally expensive, making them unsuitable for scenario analysis and real-time forecasting during emergencies.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What surrogate models are compared in the study?",{"text":80,"@type":76},"A machine learning model based on a neural network (LSTM) and a conceptual model based on HAND-style simplified hydraulic concepts are developed and compared.",{"name":82,"@type":73,"acceptedAnswer":83},"How accurate are the two approaches for predicting overland flow propagation?",{"text":84,"@type":76},"The machine learning model reaches around 95% accuracy, while the conceptual model achieves about 70–80% accuracy.","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},[93,97,101,105,110,115,119,122,127,130,134],{"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":21,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]