Is the connectivity function a good indicator of soil infiltrability distribution and runoff flow dimension?

Among the studies on runoff connectivity of soils with heterogeneous properties, the need to understand the relationships between soil heterogeneity and the associated runoff organization and amount is frequently mentioned. In this study, we simulate the stationary runoff–runon process on bi‐dimensional (2D) flat slopes for five infiltrability distributions, one of them correlated, as a function of rainfall intensity and flow dimension. We define flow dimension by 1 + ε, where ε is the outflow fraction transferred from one pixel to each of the two lateral downslope pixels. Our aim is to assess the effect of ε and soil heterogeneity on the connectivity function compared to the mean runoff flow rate, the wet area and the number of runoff patterns. The analysis of connectivity is carried within the percolation framework. The results show that the integral connectivity scale is more sensitive to the flow dimension and soil heterogeneity compared to the other variables. The wet area fraction does not depend on ε. Unlike previous studies, we find that increased runoff production is not necessarily related to increased connectivity. The use of the connectivity function within the percolation framework appears to be a valuable method for assessing the impact of soil heterogeneity and flow dimension on the runoff organization during a rainfall event. Copyright © 2014 John Wiley & Sons, Ltd.