Skip to main content
European Commission logo print header

An european flood forecasting system

Deliverables

Description: The methodology to be used is based on the Generalised Likelihood Uncertainty Estimation (GLUE) approach developed at Lancaster. GLUE (Generalised Likelihood Uncertainty Estimation) is a methodology based on Monte Carlo simulation for estimating the predictive uncertainty associated with models. A review of the approach and past applications has been published as Beven et al. (2000a). Potential Applications: The methodology is potentially applicable in all studies where numerical models are used which results depend on uncertain parameters, e.g. environmental models. End-users: Research institutes that develop environmental models (environmental in wide sense) Main innovative features/benefits: The method is relatively easy to understand and can be applied also in models where the probability distribution in space and time of the model parameters is not known. Analysis of the market or application sectors: The method can be freely used, however it is not yet commonly be part of environmental models. As increasingly complicated models, containing many difficult determinable parameters, are used for estimating environmental risks, a tool that is capable to provide a measure of uncertainty in such models can be of great interest. Potential barriers: The method is very time consuming when applied on models that contain many parameters. This makes application in models that are used short period where runtime plays major role.
Description: The computer code of the LISFLOOD model is adapted to forecast discharge for different basins. The LISFLOOD model uses grid maps for weather variables provided by the Meteorological partners as well as grid maps for elevation, land use, soil type and drainage pattern on 1 kilometre resolution or 5 kilometre resolution. The model provides at user-selected locations (grid cells) time series for discharge. The model can also provide maps of snow extent, soil moisture, and other hydrological variables. Potential Applications: Within this project the LISFLOOD model is applied for flood forecasting. The model, however, is capable to provide not only river discharges but also a range of other products such as maps showing soil moisture or snow cover extent. As it is based on raster-GIS, it can relatively simply apply using freely available raster data (e.g. USGS) as well as on raster based weather forecasts and remotely sensed information. This makes it a powerful tool that can be used not only for flood forecasting in Europe, but also for water resources planning, studying effects of climate and land-use change on water resources in different location in the world. End-users: Water Authorities, Spatial Planning Authorities, Research institutes and consultants that deal with hydrology and/or water resources planning. Main innovative features/benefits: LISFLOOD is the first distributed model that is potentially designed for continental scale flood forecasting. Although many models exist that are suited for flood forecasting, the advantage over current state of the art models applied for flood forecasting is that LISFLOOD can be linked directly to the modern grid weather forecasts, as well as to global raster data bases. It can be foreseen that not only weather forecasts but also observed weather will be provided more and more in form of maps. Hydrological models that can directly deal with such information are easier and cheaper to apply, than models that require sophisticated interpolation procedures. Analysis of the market or application sectors: The model could be applied in river basins where no forecasting system is used. Also a possible European institute for flood forecasting. The model could also serve as a backup model next to existing systems. Potential barriers: The barriers for using LISFLOOD for operational flood forecasting in specific basins is that forecasters are not used to the model. As it is a new model is not yet “proven technology”.
Description: The development of a computer program that can serve as a forecasting system is ready. This instrument integrates the weather forecasts with the hydrological models. It will allow the user to make flood forecasts in an easy way, without exact knowledge on the formats, and characteristics of the weather data, the hydrological models, the parameter uncertainty techniques etc. The instrument enables the user to view the weather forecasts, it will validate imported data and warn the user for suspicious import data, and it will allow the user to check and edit these suspicious data. The system will run the LISFLOOD model and show its results. The instrument will provide the flood forecasts made by the hydrological models for different basins, or alternatively show the local forecasts from other sources. Potential Applications: The system is a generic system that can be used for flood forecasting in all basins in Europe. End-users: Institutes that are responsible for flood forecasting. Main innovative features/benefits: The designed system has the following advantages above current systems: It assists the user to carry out the right procedures required for a forecast, which eases the flood forecasting routine - It makes directly use of modern grid weather forecast - It is able to deal with both lumped hydrological models as well as with distributed hydrological models - It allows the user to view both forecasted discharges at gauging stations as well as raster maps with spatial hydrological information - It provides both a normal forecast and added alternative forecasts based on the ensemble weather predictions. Analysis of the market or application sectors: Potential barriers:
After the recent severe floods (especially extreme flood event of July 1997 and August 2002) crisis managers realised that the nations and European-wide Hydro-meteorological Services did not cover (and cannot cover) small scale/region operational systems with extremely short lead-time for floods sufficiently. Existing operational systems and services will have to be considered/implemented. Changes will be made gradually. From partner point of view the results of the project will be devoted to understand and fulfil the real and potential end-users needs. Improvement of the flood forecasting system - cooperation with scientists and end-users (additionally for IMGW Wroclaw, Poland). As the improvements of meteorological and hydrological forecasts are done on the basis of dense observational network data, the database for selected flood situations in the Odra and the Kaczawa catchments were created. It contains several meteorological elements as well as hydrological data for many gauging stations. For the Kaczawa catchment, representative for the region of southwestern Poland, physiographical study was prepared and several parameters were calculated. New software for potential evaporation calculation was elaborated, tested and applied. The NAM model as the local hydrological model for Kaczawa catchment was calibrated and validated. To identify the needs and requirements of the end-users a questionnaire was prepared, afterwards cooperation issues were discussed during the special workshop (first on such scale). Potential applications: data - for further studies in Odra and Kaczawa catchments; potential evaporation software - for many stations; improved flood forecasting for the Kaczawa catchment - pattern for applications for other catchments in the region Benefits: thanks to the study the Kaczawa catchments is actually very well recognized. The potential evaporation software gives opportunity to quick and correct calculations. End-users: research institutions, water authorities, local authorities. Analysis of the market or application sectors: Flood Forecasting Services. Potential barriers: licence for NAM model.
For the objectives of the EFFS project the Hron river basin with an area of 1766 km2 located in Central Slovakia was chosen, as a representative basin for mountainous regions. For runoff simulations in the basin the HBV96, the Hron and the LISFLOOD models were used and compared. The models were calibrated on data from the 1990–1994 period and validated on data from the 1995-1999 period. Results of runoff simulations with all models were compared using 2 performance criteria. The applicability of all models in the pilot basin was confirmed. Flood hindcasts based on following meteorological forecasts for two historical flood events were done: ECMWF deterministic precipitation forecasts for the July 1997 and August 2002 floods, DWD GME, DWD LM, DMI Hirlam and ALADIN forecasts for the August 2002 flood. Relatively good timing of forecasted and observed flood peaks was achieved, overestimating of forecasted flood peaks using the DWD, ECMWF and DMI models and underestimating forecasted flood peaks using the ALADIN meteorological forecasts in comparison with observed floods was observed. For the Hron basin one and two day ahead flood forecasts using the DWD, DMI and ECMWF models, as so as five day ahead flood forecasts using DWD GME gave acceptable results. Results of the EFFS project will be implemented in the existing flood forecasting system of the Hron river basin. Seminar with the local end-users was held in Banska Bystrica in September 2003.

Searching for OpenAIRE data...

There was an error trying to search data from OpenAIRE

No results available