Projects / Programmes
Flood analysis using Copula functions
| Code |
Science |
Field |
Subfield |
| 2.20.00 |
Engineering sciences and technologies |
Hydrology |
|
| Code |
Science |
Field |
| P500 |
Natural sciences and mathematics |
Geophysics, physical oceanography, meteorology |
| Code |
Science |
Field |
| 2.07 |
Engineering and Technology |
Environmental engineering
|
Floods, high waters, return period, extreme events, copula, flood frequncy analysis
Researchers (8)
Organisations (1)
Abstract
Water management, planning, flood mapping, hydrologic design etc. all require a detailed knowledge of flood event characteristics. At the moment the return period of the extreme events is estimated upon frequency analyses of peak discharges as univariate statistical approach in Slovenia. That could lead to underestimation of the risk associated to a given event. As an example, an event with a peak of a 100-years return period could be less damaging than an event with a 10-years return period both in peak and volume. Complex hydrological events such as floods always appear as the consequence of few correlated random variables (peak, volume, duration, etc.). Therefore single-variable frequeny analysis can only provide limited assesment of these events. To fully understand all three variables and their relationship a multivariate statistical approach is neccesary. In the last years, some multivariate approaches were introduced in hydrological and environmental studies around the world. In the majority of these studies, two fundamental assumptions have been made. First, the flood variables have the same type of the marginal probability distribution and second, the variables follow the normal distribution. In reality, however, flood variables are dependent, do not follow, in general, the normal distribution and do not have the same type of marginal distribution. Following this observation, the concept of Copula has been used recently in flood frequency analysis. Copula function was succefully used in economy (finance) before. With Copula function we can model the dependence stucture among more variables independent of the types of marginal distributions they follow.
In the research project we will estimate the main characteristics of the flood hydrographs.(peak, volume, duration) from the measured hydrographs for the selected Sava (or other) catchment. Base flow will be separated from every single flood event. Then the dependence between all three flood characteristics variables (peak, volume, duration) will be estimated. The dependance will be estimated using Pearson's linear correlation coefficient and Kendall's coefficient of correlation. Best fitted marginal probability functions will be applied to every single variable not neccesary from the same family of distribution functions. At the end one of the Copula functions will be used to generate joint distribution function of all three variables. R statistical software (or other similar software) will be used for statistical analysis. With such a complex analyses it is possible to obtain various occurrence combinations of flood peaks, volumes and durations for a given flood event return period. It is also desirable in flood frequency analysis to obtain information concerning the occurrence probabilities of flood volumes under the condition that a given flood peak or duration occurs.
Development of new approach and its application is in progress at that moment. There are still open questions about the choice of samples, metodologies and results analyses. With the project Slovenia could join the leading countries and equally contribute to development and application of copula approach.
Significance for science
Field of waters undoubtedly belongs among one of the main public interests. Project follows EU Flood Directive (Directive 2007/60/EC), which aims to reduce and manage flood risk for human, the environment, cultural heritage and economic activity. Despite the significant progress of hydrology and also flood frequency analysis in recent decades, the methodology of the estimation of the risk of floods in the most of the countries has not been updated for a long time. Only a limited number of studies dealing with the estimation of the return period and analysis of extreme events with the multivariate approach have been reported in the hydrologic literature. Copula models are just beginning to make their way into the hydrological literature. Recently published studies show that the new approach of multivariate flood frequency analyses is very successful and necessary for contemporary and successful planning and water management. The research project is of great importance for the development of hydrology and other environmental sciences in the world, which is indicated by the publication of the project results in reputable international journals (5 SCI papers) and presentation of the project results at important international conferences (6), interest for them, responses, and already first citations. The research presents and introduces a new approach for determination of the return period of extreme events. Development and use of multivariate flood frequency models will be probably necessary In the future. In addition to the goals of the project, we performed trivariate frequency analyses with copulas also for flood peaks, volumes and suspended sediment concentrations, which is the first such attempt in the world. We also carried out the first attempt of assessing of the missing measurements (concentrations of suspended sediment concentrations) and presented results very successfully at an international scientific conference in Deggendorf, Germany. Development and application of copuła functions is in full progress. A great range of issues about the selection of samples, methodology and analysis of the results was answered within this project (see publications). With this project we joined the leading countries in developing of the new approach and add a unique contribution to the development and use of copuła functions in the science and practice.
Significance for the country
Field of waters undoubtedly belongs among one of the main public interests, therefore such a research is very important for the society. Floods are one of the most frequent and costly natural disasters. Not a small portion of Slovenian population, infrastructure, industry and even cultural heritage is located in flood risk areas. As a result, floods cause large damage, costs and sometimes also lives. Engineers cannot stop floods from occurring, but we should seek structural and nonstructural strategies to reduce the risk of social vulnerability, large economic losses, environmental damage and loss of life. Development of economically efficient and rational plans requires good estimation of the risk of the flooding and considering of EU Flood Directive (Directive 2007/60/EC). Accurate estimates of the magnitude and frequency of flood flows are needed for the design and operation of water-use and water control projects, for floodplain definition and management, for the design of transportation infrastructure such as bridges and roads etc. The results of the project bring a new approach of determining of the return period of extreme events in hydrological practice in Slovenia and abroad. In the future, the development and use of the multivariate probabilistic models would be necessary also in Slovenia. Furthermore, that would be necessary also due to the more frequent occurrence of extreme events (climate change). The study showed that, compared with conventional univariate frequency analyses, the copuła approach is a bit more complicated, but final results are different from those obtained by univariate analyses. We believe that the new approach will largely improve the determination of the risk of extreme events, mainly for the needs of construction of the important water structures and reduce flood damage and economic costs, social vulnerability, natural and cultural heritage vulnerability and probably save also lives.
Most important scientific results
Annual report
2013,
final report,
complete report on dLib.si
Most important socioeconomically and culturally relevant results
Final report,
complete report on dLib.si
