Title Procjena poplavnog rizika s utjecajem nesigurnosti : disertacija
Title (english) Flood risk assessment under uncertainty : doctoral thessis
Author Toni Kekez
Mentor Snježana Knezić (mentor)
Mentor Roko Andričević (komentor)
Committee member Ivo Andrić (predsjednik povjerenstva)
Committee member Lidija Tadić (član povjerenstva)
Committee member Goran Gjetvaj (član povjerenstva)
Granter University of Split Faculty of Civil Engineering, Architecture and Geodesy (Department of Hydrotechnical Engineering) Split
Defense date and country 2021-03-11, Croatia
Scientific / art field, discipline and subdiscipline TECHNICAL SCIENCES Civil Engineering Hydrotechnology
Universal decimal classification (UDC ) 626/627 - Hydraulic engineering and construction. Water (aquatic) structures
Abstract Mnoge zemlje diljem svijeta često su pogođene poplavama uslijed intenzivnih oborina, otapanja snijega te lomova nasipa i brana, koje rezultiraju s velikom materijalnom štetom ali i ljudskim žrtvama. Europska Direktiva o poplavama karakterizira poplavni rizik kao kombinaciju vjerojatnosti poplave i mogućih štetnih posljedica na zdravlje ljudi, okoliš, kulturnu baštinu i gospodarsku aktivnost koje se povezuju s poplavom. Upravljanje poplavnim rizikom odnosi se na ublažavanje i prevenciju negativnih posljedica poplava donošenjem adekvatnih odluka za implementaciju različitih mjera. Nesigurnost igra značajnu ulogu u upravljanju poplavnim rizikom, posebice neposredno prije i tijekom poplavnog događaja. U ovom doktorskom radu predložen je okvir za implementaciju nesigurnosti ponašanja sustava u procjeni poplavnog rizika, razvijen u svrhu poboljšanja procesa donošenja odluka tijekom odgovora na katastrofu. Nesigurnosti su kvantificirane korištenjem modela Bayesove mreže te je razvijeni okvir validiran na području Slavonskog Broda gdje se pokazalo da bez obzira na poboljšanje sustava obrane od poplava još uvijek postoji značajna opasnost uslijed mogućeg otkazivanja ustave koja štiti naseljeno područje. Također, kao rezultat istraživanja pokazalo se da bez obzira na mogućnost otkazivanja ustave, pravovremenom evakuacijom stanovništva veličina poplavnog rizika se može značajno smanjiti. Strateško planiranje prilikom upravljanja poplavnim rizikom bazira se na kvantifikaciji poplavnog rizika kroz izračun očekivane godišnje poplavne štete. U ovom doktorskom radu prezentirana je polu-analitička metoda za procjenu očekivane godišnje štete temeljena na faktor graf modelu, koja omogućuje obuhvaćanje cijelog prostora vjerojatnosti kao i fleksibilnost prilikom definiranja ulaznih parametara. Nadalje, uveden je novi način definiranja distribucije godišnje štete izračunom prva dva statistička momenta te njihovom implementacijom kroz Beta distribuciju. Analizom distribucije godišnje štete i različitih utjecaja nesigurnosti, rezultati su pokazali da postoji značajan utjecaj ekstremnih događaja malih vjerojatnosti na veličinu očekivane godišnje štete.
Abstract (english) Many river basins are experiencing frequent flooding events with significant economic and other losses due to intensive precipitation as well as other atmospheric and hydrological conditions. European Flood Directive defined flood risk as a combination of flooding probability and possible adverse consequences on people, assets, cultural heritage and environment. Flood risk management considers implementation of different measures for mitigation and prevention of possible negative consequences related to flooding. Uncertainty can strongly affect the flood risk management process, especially near and during the flood event. A framework is proposed for implementation of uncertainty related to behavior of the endangered system in the flood risk assessment, in order to improve the decision-making process during the flood emergency response. The proposed framework is validated on the City of Slavonski Brod pilot site, where the results demonstrated that there is a significant flood hazard still present due to possible weir failure, despite the improvement of flood defense measures. Furthermore, the results demonstrated how flood risk value can significantly decrease by properly evacuating the affected population. Flood risk management on a strategic level requires a monetary quantification of possible flood risk, which is performed by calculating expected annual damage (EAD) based on the combination of flooding event probability and corresponding damages. A semi-analytic methodology is presented for estimation of expected annual damage based on the factor graph model, which enables integration of entire probability space as well as flexibility in defining input data. Furthermore, a novel approach is presented for definition of annual damage distribution based on first and second statistical moment and by employing Beta distribution. By analyzing the annual damage distribution as well as impact of different sources of uncertainty, the results demonstrated that there is a significant impact of extreme events with low occurrence probability on the expected annual damage.
Keywords
poplave
poplavni rizik
nesigurnost
Bayesove mreže
očekivana godišnja šteta
faktor grafovi
distribucija godišnje štete
Keywords (english)
floods
flood risk
uncertainty
Bayesian networks
expected annual damage
factor graphs
annual damage distribution
Language croatian
DOI https://doi.org/10.31534/DocT.053.KekT
URN:NBN urn:nbn:hr:123:862199
Promotion 2021
Study programme Title: Civil Engineering Study programme type: university Study level: postgraduate Academic / professional title: doktor/doktorica znanosti, područje tehničkih znanosti, polje građevinarstvo (doktor/doktorica znanosti, područje tehničkih znanosti, polje građevinarstvo)
Catalog URL http://library.foi.hr/lib/knjiga.php?B=422&H=&E=&V=&lok=&zbi=&item=14941
Type of resource Text
Extent X, 99 listova : graf. prikazi, ilustr. u bojama ; 30 cm
File origin Born digital
Access conditions Access restricted to students and staff of home institution
Terms of use
Created on 2021-04-12 13:48:02