Flood simulation and prediction

profilová_Vědy o zemi

The Floreon+ system uses rainfall-runoff models to simulate and predict river flow rates based on the measured and predicted total precipitation, temperature and snow height and water equivalent. This information is then used to run hydrodynamic models in order to predict and visualise the area that may be flooded during a critical situation. The extent of floods can thus be predicted several hours or even days before such a situation actually occurs, and it is therefore possible to prepare for the coming floods in time, minimizing both risk to human life and financial losses. Since the input data for these models (especially total precipitation predictions) may contain a certain degree of uncertainty, it is also necessary to include multiple levels of these uncertain data in the calculations and provide information about all likely scenarios. Besides crisis situations, the Floreon+ system allows for conducting WHAT-IF analyses that may for example reveal the impacts of modifications to river channels and surrounding areas, identify floodplains and determine the effectiveness of flood control measures.


Simulations of torrential rainfall and surface runoff


It has been well documented that the phenomenon of torrential rainfall and flash floods occurs in the Czech Republic with an increasing intensity. The approach that is applied within the Floreon+ system is based on the use of fully distributed models where the territory is represented through a computational network. The computation is performed within this computational network and the outputs – such as surface runoff – can be immediately visualized. Therefore, this technology can be used to produce flash flood predictions sufficient time in advance (hours). Even though, in meteorological terms, the prediction of torrential (convective) precipitation is a complex phenomenon, geoinformatic analyses can be used to create maps of the territory’s vulnerability to flash floods due to torrential rainfall. Subsequently, in the event torrential rainfall actually occurs, the bodies that are involved in crisis management have a set of geographical layers or maps at their disposal that allow for forecasting the risks existing within the territory affected by the rainfall.


The calibration and verification of rainfall-runoff models


The calibration of rainfall-runoff models can certainly be viewed as the basic tool for providing reliable, high-quality output from such models. Therefore, the Floreon+ research team addresses the identification of selected parameters of rainfall-runoff models that are burdened with uncertainty using inverse modelling tools. The main objective is to develop a method that would clearly identify the values of model parameters for selected rainfall-runoff events that would provide the most accurate outputs. It is also important to be able to validate the rainfall-runoff models that are used and to evaluate their results, both for the possibility of comparing them and identifying the best models for different situations, and for the indication of the need to further calibrate their parameters. The Floreon+ system uses both graphical and basic statistical evaluation to compare the models’ results with actual measured values and uses them to evaluate these models.


Air pollution dispersion modelling


Air quality is a very closely monitored factor in the Moravian-Silesian Region. The large concentrations of pollutants result from the large concentration of industrial enterprises, the ever increasing traffic density and the large number of single-family houses that use solid fuels for heating. In addition, the above factors are further exacerbated by the relatively frequent occurrence of inversions. The deployment of these models within the Floreon+ system will allow for obtaining a short-term forecast of air pollution concentrations due to increased emissions from sources that include industrial enterprises, local heating and current traffic situation. Also, the current meteorological situation is an important factor contributing to increased air pollution. In particular, great emphasis will be placed on the prediction of inversion situations. All these outputs of simulations should help timely predict increased air pollution concentrations at certain locations and, in turn, adopt necessary measures to improve the situation, such as traffic restrictions, reduced production etc.


Modelling leakages of hazardous substances


Deploying this module within the Floreon+ system should mainly be useful in crisis management. Floreon+ should provide predictions of hazardous substance dispersion due to accidents. It is assumed that the leakage of hazardous substances may occur at a predetermined location with previously known contaminants (the storage of hazardous substances within the premises of industrial enterprises) or possibly from a mobile source, typically the road or rail transport of hazardous substances. As opposed to emergency plans, the Floreon+ system will also take into account the current meteorological situation. Such refined prediction will then allow for a more efficient evacuation of the affected area and, as the case may be, liquidation of the accident’s consequences.

PRACE Summer of HPC 2018 has launched
Early-stage postgraduate and late-stage undergraduate students are invited to apply for the PRACE Summer of […]
Invitation to the course Intel Xeon Phi programming
(2018-02-22 to 2018-02-23)
When: Thursday February 22, 2018, 9.30am – Friday February 23, 2018, 4.30pm Where: campus VŠB-TUO […]
European Commission proposes to invest in world-class European supercomputers
The European Commission unveiled on 11 January 2018 its plans to invest jointly with the […]
The Czech Minister of Industry and Trade, Tomáš Hüner, visited IT4Innovations
During his first business trip in office, the new Minister of Industry and Trade Tomáš […]
IT4Innovations Newsletter Q4/2017
IT4Innovations Newsletter Q4/2017
All news