Godderis(eの頭に´) et al.(2006)による〔『Modelling weathering processes at the catchment scale: The WITCH numerical model』(1128p)から〕


 A numerical model of chemical weathering in soil horizons and underlying bedrock (WITCH) has been coupled to a numerical model of water and carbon cycles in forest ecosystems (ASPECTS) to simulate the concentration of major species within the soil horizons and the stream of the Strengbach granitic watershed, located in the Vosges Mountains (France). For the first time, simulations of solute concentrations in soil layers and in the catchment river have been performed on a seasonal basis. The model is able to reproduce the concentrations of most major species within the soil horizons, as well as catching the first-order seasonal fluctuations of aqueous calcium, magnesium and silica concentrations. However, the WITCH model underestimates concentrations of Mg2+ and silica at the spring of the catchment stream, and significantly underestimates Ca2+ concentration. The deficit in calculated calcium can be compensated for by dissolution of trace apatite disseminated in the bedrock. However, the resulting increased Ca2+ release yields important smectite precipitation in the deepest model layer (in contact with the bedrock) and subsequent removal of large amount of silica and magnesium from solution. In contrast, the model accurately accounts for the concentrations of major species (Ca, Mg and silica) measured in the catchment stream when precipitation of clay minerals is not allowed. The model underestimation of Mg2+ and H4SiO4 concentrations when precipitation of well crystallized smectites is allowed strongly suggests that precipitation of well crystallized clay minerals is overestimated and that more soluble poorly crystallized and amorphous materials may be forming. In agreement with observations on other watersheds draining granitic rocks, this study indicates that highly soluble trace calcic phases control the aqueous calcium budget in the Strengbach watershed.

1. Introduction
2. Model description
 2.1. The WITCH model
 2.2. Coupling with a model of water and carbon cycles in forest ecosystems
 2.3. Sites, run design, forcing functions and parameters
3. Results of the reference simulation
 3.1. Reference run, PP location
 3.2. Reference run, HP location
 3.3. Calculating the chemical composition of the main stream at the spring collector
4. AP and APL10 sensitivity tests: Looking for calcium at the spring collector
5. NOPREC and KSP simulation: Looking for silica at the spring collector
6. Trace mineral and base cation budget of a granitic catchment
7. Some uncertainties and limitations
8. Conclusions

Table 5 Contribution of trace minerals to the total base cation flux exported from the watershed
  Main stream BC flux (weathering contribution only)
Percentage of the Ca flux originating from trace mineral dissolution relative to the total BC flux
Hubbard Br. 101 (Drever and Clow, 1995) 23.5% (trace apatite) (Blum et al., 2002)
Estibere(最初のeの頭に`) 167 (Oliva et al., 2004) 80% (trace silicates) (Oliva et al., 2004)
Loch Vale 42 (Mast et al., 1990) 41% (trace calcite) (Clow et al., 1997; White et al., 1999)
Strengbach Measured: 281 15% (trace apatite) (Aubert et al., 2001)
KSP simulation (spring collector): 285 26% (HP) to 69% (PP) (trace apatite) (This study)