Are Water Table Variations in a Shallow Forest Soil Consistent with the TOPMODEL Concept?

Abstract

The TOPMODEL concept provides a simple method for parameterizing the effect of topography on the distribution of subsurface moisture and runoff production, and has been incorporated into models of catchment hydrology and forest ecology. Most evaluations of the TOPMODEL concept have been based on comparisons of stream flow hydrographs and do not necessarily provide a test of the predictions of subsurface flow and saturated source area dynamics. The objective of this study was to evaluate whether water table predictions based on the TOPMODEL concept agree with observed water table depths for a shallow forest soil in a small subcatchment drained by an ephemeral stream. Observed water table depths were fitted well by a linear statistical model that is mathematically consistent with the TOPMODEL assumptions and that specifies time and location effects independently. That is, the water table does not change shape but shifts uniformly up or down in response to changes in saturated zone storage. Residuals from the fitted model were generally less than 10 cm in absolute value, compared to observed ranges of water table depths up to 100 cm. The TOPMODEL concept predicts that the relation between the estimated location effects for the wells and the values of In (a/tan β) should be linear, where a is the drainage area per unit contour length, and tan β is the surface slope. Regression analysis revealed that the fitted linear relation was statistically significant (P=0.001), but weak (R2= 0.26). This weak fit could be caused by (1) significant spatial variability in soil transmissivity (assumed negligible in many applications of TOPMODEL) and/or (2) errors in specifying In (a/tan β) from a digital elevation model of the surface topography. Considering the increasingly common application of the TOPMODEL concept in hydrologic and ecological models, further research should be carried out to specify more clearly the bounds of applicability of its underlying assumptions and to identify workable extensions and modifications to the concept, where necessary.