Calibration and Validation of The Soil Water Balance Model Wave For Forest Stands In Flanders: 2. Methods and Preliminary Results
Abstract
This research, stimulated by government policy, is conducted in de framework of the reforestation of agricultural land and its impact on hydrological basins. The integra- tion of the ongoing research efforts in Flanders offers a unique opportunity to calibrate and validate the water balance module of the WAVE-model (Water and Agrochemicals in soil, crop and Vadose Environment; Van Clooster et al., 1994) for the most impor- tant soil and forest types. Hence, the scientific objectives of this project are multiple. This poster will show the first results of the calibrated and validated WAVE-model and the simulated evapotranspiration for forest stands and agricultural land use. Fur- thermore comparison will be made with sap-flow measurements using the Heat Field Deformation Method (Cermàk and Nadezhdina, 1998). The calibration and validation procedure consist of a sensitivity analysis of the model parameters, model calibration using simulated and measured time series of soil water content at different depths of the soil profile (March 2000-August 2001) and multi-site validation (using the cali- brated parameters from one site on other experimental sites). The crop factor (Kc), the saturated hydraulic conductivity (Ksat) and the root distribution function are the major calibration parameters. The Root Mean Square Error (RMSE), Model efficiency (ME) and Coefficient of Determination (CD) are the simulation statistics implemented during the calibration step. The preliminary results show time series of soil water con- tent (in cm3.cm-3 for the whole soil profile) with a RMSE of 2.97 % (optimal as small as possible), a ME of 0.58 (optimal 1.00) and a CD of 1.01 (optimal 1.00) for a beech stand at a validation site. In case of the popular stand the actual transpiration (Tact), simulated with WAVE is overestimated compared with the sap-flow measure- ments (June-September 2000). The simulation results show more peaks and fall to zero when heavily daily rain showers occur, while the sap-flow measurements indi- cate non-zero tree transpiration. The same results are found in case of the pine stand (June-September 2002) but there is only a slight overestimation and the simulated and measured time series of Tact are much more compatible.
- Publication:
-
EGS General Assembly Conference Abstracts
- Pub Date:
- 2002
- Bibcode:
- 2002EGSGA..27..547V