1. General Model Information

Name: CATFLOW a new physically based, distributed model for the dynamics of

Acronym: CATFLOW

Contact:

Author(s):

Abstract:

by Thomas Maurer

A new physically based, distributed model (named CATFLOW) for the dynamics of water in a small rural catchment on the event and season time scale is presented. Results are shown for the 3.5 sqkm research catchment Weiherbach in SW-Germany.

The model is based on the idea of subdividing the 3D-landscape in patches of 2D-hillslopes of variable width (assuming water flowing normal to the topographical contours) connected to a detailed (partly ephemeral) drainage network. Though there have been developments based on this or similar ideas in the past a new effort has been undertaken to improve concept and structure of this type of model.

With increasing size of a catchment the shape of a hydrograph is more and more influenced by the properties of the drainage network rather then by overland flow; the latter then mainly has to provide realistic estimates of event based runoff coefficients. Increasing size is a relative measure. If we trace down the drainage network into every single swale of the landscape these considerations are valid for even very small catchments.

Thus we get a partial decoupling of runoff production (at the hillslope) and runoff propagation (in the detailed drainage network), achieving a reduction of the dimension of the problem.

Runoff coefficients are determined by proper evaluation of infiltration and evapotranspiration, both at times controlled by either soil or atmospheric conditions. CATFLOW spends considerable efforts in a more precise description of these processes.

For infiltration these are:

• Consideration of macropores as introduced e.g. in the model HILLFLOW thus annulling the potential concept for the infiltration process.
• Consideration of subsurface stormflow
• Consideration of interaction between the drainage networks water level and the wetness condition at the lower end of a hillslope, strongly governing the extension of the socalled contributing areas.
• Consideration of runoff contribution from sealed surfaces.

• Relational data management for efficient and flexible input of driving atmospheric and landuse boundary conditions.
• Rootzone water uptake is controlled by a minimum resistance principle, i.e. water is always taken from that depth within the rootzone, where total potential difference to the surface is minimal.
• Dynamic development of plant parameters in the growing season is accounted for by use of plant development curves which are modulated by a quality measure for the plants location.

• Pre- and postprocessing is supported by ARC/INFO based developments to comfortably assists in spatial analysis of the catchments geometry and attributes.
• Hillslope geometry is described by use of orthogonal, curvilinear coordinates to (a) flexible model the irregular boundaries and (b) to more easily conduct numerical solution of Richards equation for matrix flow on a mathematical rectangular grid.
• Potential based solution of Richards equation using Picard iteration and Taylorseries expansion of the moisture content with respect to the potential.
• Soil hydraulic properties are provided in tables, thus flexible to different data sources. The model allows to account for anisotrophy.
• Channelflow within the drainage network is calculated based on St. Venant equations or Muskingum-Cunge method.
• Each hillslope has its individual time step control, thus reducing overall CPU time consumption.
• Development of similarity criterions for hillslope behaviour are under examination, thus further reduction of CPU time consumption might be possible.

Source of the abstract:
CATFLOW Home-Page

II. Technical Information

II.1 Executables:

II.2 Source-code:

II.3 Manuals:

II.4 Data:

III. Mathematical Information

III.1 Mathematics

III.2 Quantities

III.2.1 Input

III.2.2 Output

IV. References

V. Further information in the World-Wide-Web

• Thesis of Thomas Maurer about CATFLOW

VI. Additional remarks

Go back to Register of Ecological Models (R E M)