Comment on “Effect of Rainfall‐Induced Soil Seals on Soil Water Regime: Wetting Processes” by Y. Mualem, S. Assouline, and D. Eltahan

Abstract

In their paper, Mualem et al. [1993] have provided an interesting exposition of the effects of a surface layer on infiltration and wetting profiles resulting therefrom. The results confirm our understanding of the effect of surface soil layers [e.g., Whisler et al., 1972;Miller and Gardner, 1962], and are consistent with the general theory for infiltration through a surface layer presented earlier [Smith, 1990]. However, one apparent result suggested in this manuscript is misleading, which I wish to clarify here.In Figure 1, Mualem et al. [1993] show a simulated infiltration pattern in which surface infiltration rate q0 is plotted as a function of cumulative rainfall, R, for two soils.In the abstract and the subsequent discussion, it is suggested that the presence of a seal or surface layer would cause the final infiltration rate, qf, to be larger for a larger application rate I (the symbol convention of the authors is adopted here). In their equation (1), qf is used conventionally as an asymptotic rate, but in the discussion (p. 1655) this "final rate" is arbitrarily defined as the rate after 50 mm of rainfall. Normally, final rate is an asymptotic rate, and this distinction is significant in the points to be made here. For this discussion I define the asymptotic rate as q•. The theoretical value of q• is a function only of the properties of the soil, including the relative properties of the surface layer and the underlying soil if there is a surface layer [Smith, 1990]. The suggestion that the "final" infiltration rate is a function of rainfall I is misleading and apparently a result of the choice to plot q0 against cumulative rainfall, R. After ponding, for as long as I > q0, q0 is only controlled by conditions within the profile, and is independent of I or R. This is inherent in the definition of ponding. By choosing to plot q0(R), the authors inadvertently distort each curve proportional to the value of I. The infiltration rate occurring after a certain depth of rainfall, qf, will always be larger for larger constant rainfall rates, regardless of the presence or absence of a surface seal. Thus their discussion in this regard is misleading with respect to the effect of a