OPAQUE -

Operationelle Abfluss- und Hochwasservorhersage in Quellgebieten   operational discharge and flooding predictions in head catchments

OPAQUE Methods

The water balance model LARSIM

Fig. 1: Scheme of the water balance model LARSIM

Water balance models are mathematical computation procedures to continuously describe and quantify the spatial and temporal distribution of the water balance's essential components like precipitation, evapotranspiration, infiltration, water storage and runoff. The water balance model LARSIM (Large Area Runoff Simulation Model) is a so-called conceptional model, i.e. the complex processes in the natural system are reproduced by simplified model concepts.

After a model-intern regionalisation and, if necessary, correction of meteorological parameters in LARSIM among others the following hydrological processes are considered: interception, evapotranspiration, snow accumulation, snow compaction and snow melt, soil water storage as well as storage and lateral transport in streams and lakes (LARSIM model scheme, figure 1).

Figure 2: System data for the LARSIM water balance models in Baden-Wuerttemberg

Input data

Hydrometeorological data

Meteorological model input data are measured or by use of climatic models computed time series for precipitation, air temperature, relative air humidity, wind speed, global radiation and air pressure. Concerning hydrological input data measured gauge runoff, water transfer and water management procedures like dams, regulation of lakes and retention basins can be considered. If LARSIM is applied for operational flood forecasting furthermore measured snow height or specifications of snow water equivalent can be used for an automatical optimization of the LARSIM snow module.

System data

Specifications of ground level, slope, land use, field capacity, stream geometry and the location of water level gauges and climatological gauges are needed as system input data (figure 2). These informations are largely extracted computer-aided by use of geographical information systems.

Figure 3: Processing of LARSIM system data in a GIS exemplary for the data sets ground level, land use and stream network

For each single raster element up to 16 different land use categories with specific evapotranspiration and runoff characteristics are distinguished (figure 3). By use of ArcView-interfaces land use scenarios can be defined to evaluate their impact on water balance.

Figure 4: Example for the digital river network (top) and the river routing scheme of the model (bottom) (grid size 1x1 km)

The reproduction of the real stream network (raster elements' crosslinking ) is done by a calculative intersection of the digital stream network and the model raster (figure 4). For each stream section geometric specifications about stream length and decline as well as width and height of the mean cross section are incorporated in the LARSIM system dataset.

Based on the meteorological input data the water balance model computes spatially and temporally highly resolved states of the terrestrial water cycle's components (like evapotranspiration, soil moisture and runoff).

If LARSIM is used for operational runoff forecasting an automatical optimization routine analyses the difference between measured and calculated runoff and performs a correction of water yield or the actual water storage content according to the respective runoff situation. Based on the optimised model state the runoff forecasts are computed (figure 5). Also the model intern snow cover can be automatically adapted to measured snow data.

Figure 5: Example of a flood early warning computed with LARSIM (gauge Schwaibach/Kinzig in the Black Forest, basin area 954 km²)

The model results can also be exported and visualised temporally aggregated as spatial values, for example as maps of actual soil moisture or height of snow cover (figures 6 and 7).

Figure 6: Exemplary map of relative soil moisture in Baden-Wuerttemberg (approx. 36.000 km²) calculated with LARSIM

Figure 7: Exemplary map of snow cover height in Baden-Wuerttemberg (approx. 36.000 km²) calculated with LARSIM

The water balance model LARSIM can be operated in different temporal and spatial resolutions and is therefore suitable for many different tasks. Some examples are:

• Continuous forecast for low, mean and high water in an operational mode
• Forecast of water temperature
• Estimation of impact of environmental change like climate change or land use changes on water balance
• Prognosis and scenarios for river development planning
• Supra-regional determination of ground water recharge as basis for sustainable economy
• Supply of area-wide infiltration and runoff data for water quality models

The model LARSIM is used in the flood forecast centre Baden-Wuerttemberg and other flood warning centres, among other things, for operational flood and low water forecast as well as forecast of water temperature. Detailed information about the model basics and applications on different spatial scales can be found in Ludwig & Bremicker (2006), Bremicker (2000), Bremicker et al. (2006) und Luce et al. (2006).

Literature

Ludwig, K., Bremicker, M. (Eds.) (2007): The Water Balance Model LARSIM. Freiburger Schriften zur Hydrologie, Band 22 Institut für Hydrologie der Universität Freiburg. (),

Bremicker, M. (2000): Das Wasserhaushaltsmodell LARSIM - Modellgrundlagen und Anwendungsbeispiele. Freiburger Schriften zur Hydrologie, Band 11 Institut für Hydrologie der Universität Freiburg. (),

Bremicker, M., Homagk, P., Ludwig, K. (2006): Hochwasserfrühwarnung und Hochwasservorhersage in Baden-Württemberg. Wasserwirtschaft, 7/8 , S. 46-50

Luce, A., Haag, I., Bremicker, M. (2006): Einsatz von Wasserhaushaltsmodellen zur kontinuierlichen Abflussvorhersage in Baden-Württemberg. Hydrologie und Wasserbewirtschaftung, 50 (2), S. 58-66