| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department for Agro-Forestry Systems Management, Hydraulics Division, University of Basilicata, Potenza, Italy
acoppola{at}unibas.it
Crescimanno and Garofalo (2005) simulated water flow and solute transport in a cracking soil irrigated with saline water by using the SWAP model (van Dam et al., 1997). The SWAP model provides as outputs, among others, the water content as well as the electrical conductivity. The authors compared the simulated results with measured values of soil moisture and electrical conductivity of the saturated extract (ECsat).
The crucial issue I want to emphasize is that the authors derived the simulated ECsat by using the convection–dispersion equation for a conservative solute (Eq. [4] in the paper). Instead, the measured ECsat results from several physicochemical processes, most of them (such as pure retardation, sorption/desorption, degradation, etc.) are quite important, and they were not accounted for by the authors. Such phenomena are fundamental processes that influence to a large extent ion mobility as well as their partition between sorbed and solution phases (e.g., see the numerous contributions of van Genuchten and coworkers). For instance, neglecting retardation implies an underestimation, which can be also considerable depending on solution components, pore velocity, and dispersion.
In essence, the authors have simulated transport of a conservative solute, which is not in compliance with measured values. More important, my main criticism is that the model (once again the convection–dispersion equation for conservative solutes) they have used is intrinsically mass conservative whereas their ECsat measurements refer to solutes, which to the contrary may undergo degradation as well as sorption/desorption on the soil. Summarizing, the authors have used a propagation model, which is not representative of the real physicochemical processes from which their measurements arise.
I wonder about the rationale behind the choice of using a conservative form of convection-dispersion equation, since the SWAP version the authors have quoted in the paper allows for convective–dispersive transport coupled with adsorption, and decay (see Eq. [3.14], page 43 in van Dam et al., 1997). Notwithstanding this option, Crescimanno and Garofalo (2005) state on p. 1945: "Another simplifying assumption concerning solute transport in SWAP is that no distinction is made between different cations and anions, and only the total amount of salts is considered."
In conclusion, the parameters estimated by the authors must be considered just fitting parameters. The relatively good agreement between measured and simulated ECsat (Fig. 6, p. 1952) is largely expected in a calibration context. Given the fitted parameters, independent ECsat data would be reproducible (in the sense that they have not been used in the calibration procedure). This is the only support provided for their contention, "SWAP predicts solute concentration ...neglecting the process of cationic exchange. Due to these as well as to other simplifying assumptions, the accuracy of the predicted ECsat in clay soils needs to be checked by comparisons with measured ECsat values" (Crescimanno and Garofalo, 2005, p. 1944).
Besides, it should be considered that in the particular experimental conditions considered by the authors (vertic clay soils, drying and wetting cycles under arid conditions), cation-exchange processes would be especially relevant for controlling soil solution composition and the mobility of different ions. In this sense, it might be maintained that the SWAP model is misused, since it ignores chemical equilibria among cations.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
