Biogeochemical Models
299
ated at forest maturity because element retention tends to approach element
mineralization. Atmospheric input fluxes and uptake and runoff output
fluxes are introduced in the model based on the information of biogeochemical studies at the catchments of interest. The weathering rates are adjusted through an iterative process of approximation until an acceptable fit is
attained between model simulated versus observed streamwater chemistry.
Once the data set containing these parameters has been adjusted, the
model solves the chemistry of the streamwater from the system of equilibrium equations coupled with the mass balance. The model's time step is 1
year. In the following simulations the runs extended from 140 years ago up to
the present day and forecasted a variety of scenarios into the next 140 years.
Based on data from the period 1983-1988 for Montseny and 1981-1988 for
Prades, the model has been calibrated at both sites for year 1985, which has
been taken as the present year.
21.3 Application of MAGIC to Montseny
In Montseny, the TM9 catchment at La Castanya (Chap. 2) was used for calibration of MAGIC. This catchment had been monitored for several years
since 1983, with a weekly sampling schedule of bulk deposition, soil solution
and streamwater, while information on soil characteristics was readily available (Chaps. 2 and 17). The two-box version of the model was used in an attempt to simulate the water sources contributing to the stream. One box represented the groundwater end-member which generated the baseflow and
the other box represented the soilwater end-member. The relative proportions of the groundwater and soilwater end-members were determined by
simple mixing relationships between the average composition of three conservative elements (Na+, sol- and CI-) in the streamwater, soilwater and
groundwater (Avila et al. 1995).
The model was calibrated with: (1) fixed parameters: soil characteristics
(soil depth, porosity, bulk density, sulphate adsorption), soil chemistry
(exchangeable cations, base saturation), hydrological characteristics (mean
precipitation, mean runoff), biological uptake, rain chemistry and dry deposition coefficients, and (2) adjustable variables: the weathering rates. Precipitation and throughfall chemistry at Montseny from Roda et al. (1990)
provided estimates of the inputs from the atmosphere (Table 21.1). Precipitation chemistry for the initial year of 1845 was calculated by subtraction
from the current precipitation chemistry of the pollutant increment assumed
for the industrial period. The holm oak forest at La Castanya was heavily
coppiced for charcoal production until the 1950s and currently is accreting
biomass. Its net nutrient uptake was assessed from the net tree production at
La Castanya (boles and branches> 3-4 cm in diameter) multiplied by the
mean nutrient concentrations in these plant fractions, with data from Ferres
299
ated at forest maturity because element retention tends to approach element
mineralization. Atmospheric input fluxes and uptake and runoff output
fluxes are introduced in the model based on the information of biogeochemical studies at the catchments of interest. The weathering rates are adjusted through an iterative process of approximation until an acceptable fit is
attained between model simulated versus observed streamwater chemistry.
Once the data set containing these parameters has been adjusted, the
model solves the chemistry of the streamwater from the system of equilibrium equations coupled with the mass balance. The model's time step is 1
year. In the following simulations the runs extended from 140 years ago up to
the present day and forecasted a variety of scenarios into the next 140 years.
Based on data from the period 1983-1988 for Montseny and 1981-1988 for
Prades, the model has been calibrated at both sites for year 1985, which has
been taken as the present year.
21.3 Application of MAGIC to Montseny
In Montseny, the TM9 catchment at La Castanya (Chap. 2) was used for calibration of MAGIC. This catchment had been monitored for several years
since 1983, with a weekly sampling schedule of bulk deposition, soil solution
and streamwater, while information on soil characteristics was readily available (Chaps. 2 and 17). The two-box version of the model was used in an attempt to simulate the water sources contributing to the stream. One box represented the groundwater end-member which generated the baseflow and
the other box represented the soilwater end-member. The relative proportions of the groundwater and soilwater end-members were determined by
simple mixing relationships between the average composition of three conservative elements (Na+, sol- and CI-) in the streamwater, soilwater and
groundwater (Avila et al. 1995).
The model was calibrated with: (1) fixed parameters: soil characteristics
(soil depth, porosity, bulk density, sulphate adsorption), soil chemistry
(exchangeable cations, base saturation), hydrological characteristics (mean
precipitation, mean runoff), biological uptake, rain chemistry and dry deposition coefficients, and (2) adjustable variables: the weathering rates. Precipitation and throughfall chemistry at Montseny from Roda et al. (1990)
provided estimates of the inputs from the atmosphere (Table 21.1). Precipitation chemistry for the initial year of 1845 was calculated by subtraction
from the current precipitation chemistry of the pollutant increment assumed
for the industrial period. The holm oak forest at La Castanya was heavily
coppiced for charcoal production until the 1950s and currently is accreting
biomass. Its net nutrient uptake was assessed from the net tree production at
La Castanya (boles and branches> 3-4 cm in diameter) multiplied by the
mean nutrient concentrations in these plant fractions, with data from Ferres