Alternative analysis of transient infiltration experiment to estimate soil water repellency

Abstract

The repellency index (RI) defined as the adjusted ratio between soil-ethanol, S-e, and soil-water, S-w, sorptivities estimated from minidisk infiltrometer experiments has been used instead of the widely used water drop penetration time and molarity of ethanol drop tests to assess soil water repellency. However, sorptivity calculated by the usual early-time infiltration equation may be overestimated as the effects of gravity and lateral capillary are neglected. With the aim to establish the best applicative procedure to assess RI, different approaches to estimate S-e and S-w were compared that make use of both the early-time infiltration equation (namely, the 1min, S1, and the short-time linearization approaches), and the two-term axisymmetric infiltration equation, valid for early to intermediate times (namely, the cumulative linearization and differentiated linearization approaches). The dataset included 85 minidisk infiltrometer tests conducted in three sites in Italy and Spain under different vegetation habitats (forest of Pinus pinaster and Pinus halepensis, burned pine forest, and annual grasses), soil horizons (organic and mineral), postfire treatments, and initial soil water contents. The S1 approach was inapplicable in 42% of experiments as water infiltration did not start in the first minute. The short-time linearization approach yielded a systematic overestimation of S-e and S-w that resulted in an overestimation of RI by a factor of 1.57 and 1.23 as compared with the cumulative linearization and differentiated linearization approaches. A new repellency index, RIs, was proposed as the ratio between the slopes of the linearized data for the wettable and hydrophobic stages obtained by a single water infiltration test. For the experimental conditions considered, RIs was significantly correlated with RI and WDPT. Compared with RI, RIs includes information on both soil sorptivity and hydraulic conductivity and, therefore, it can be considered more physically linked to the hydrological processes affected by soil water repellency.