89th American Meteorological Society Annual Meeting

Tuesday, 13 January 2009: 2:00 PM
Application of the TOPKAPI model within the DMIP 2 project
Room 127BC (Phoenix Convention Center)
Gabriele Coccia, University of Bologna, Bologna, Italy; and C. Mazzetti, E. A. Ortiz, and E. Todini
Poster PDF (1.9 MB)
The paper presents the TOPKAPI model and its application on the Sierra Nevada basins (North Fork American Basin and East Fork Carson Basin), in relation to the Distributed Model Inter-comparison Project 2 (DMIP 2).

The TOPKAPI model is based on the idea of combining the Kinematic approach and the topography of the basin; the latter is described by a Digital Elevation Model (DEM), which subdivides the application domain by means of squared cells. Each cell of the DEM is assigned a value for each of the physical parameters represented in the model.

The integration in space of the non-linear Kinematic wave equations representing subsurface flow, overland flow and channel flow results in three ‘structurally-similar' zero-dimensional non-linear reservoir equations. The integration of the fundamental equations is performed on the individual cell of the DEM.

Beside subsurface, overland and channel flow TOPKAPI includes components representing infiltration, percolation, evapo-transpiration and snowmelt, plus a lake/reservoir component and a parabolic routing component. For the deep aquifer flow, the model accounts for water percolation towards the deeper subsoil layers even though it does not contribute to the discharge.

As precipitation falls on the catchment, the snow accumulation and melting component identifies the amount of water that actually reaches the soil surface, the snow accumulation and melting component is based on an energy budget method. The amount of water that reaches the soil surface infiltrates unless the soil is already saturated or impervious.

The soil water component is affected by subsurface flow in a horizontal direction defined as drainage; drainage occurs in a surface soil layer, with limited thickness and with high hydraulic conductivity due to its macro porosity. The surface water component is activated on the basis of an excess of infiltration mechanism. Lastly, both components contribute to feeding the drainage network. Evapo-transpiration is taken into account as water loss, subtracted from the soil's water balance.

The channel network is assumed to be tree shaped with reaches having rectangular or triangular cross sections. Channel routing is performed using a Kinematic approach for steep cells. Where channel slope is too small to use a Kinematic approach TOPKAPI uses a parabolic routing component based on Muskingum-Cunge method.

At the smallest spatial scale, the pixel element, evolution of all the hydrological state variables can be obtained: rainfall, temperature, evapo-transpiration, soil moisture conditions, snow accumulation and runoff generation. Analysis of these dynamics is extremely important for verifying the general physical soundness of the model performance as well as for calibrating parameters of certain hydrological processes.

The TOPKAPI model was applied to the Sierra Nevada basins during a period of 10 years, from the 1987 to the 1997. The model gives good results in terms of discharge and snow accumulation and melting; the latter is better reproduced in the Carson River basin because the mean elevation is high and the uncertainty about the real status of the precipitation (rain or snow) is low. It has to be noticed that in this basin, the calibration at the basin closing section was the best one even for the upstream section.

The American River basin mean elevation is lower than the Carson River's and in winter the temperature is often close to 0° C, so the uncertainty about the real status of the precipitation is high and some rain events are mistaken for snow events and vice versa. Nevertheless, due to the low elevation of the basin, the snow accumulation and melting component is not very significant, hence the TOPKAPI model well reproduces the flood events for the whole period of simulation and the evaluation indexes are optimal.

Supplementary URL: