Observation and simulation of water movement and runoff in a coarse texture water repellent soil

Water repellent or non-wetting soils show poor water infiltration, which leads to an increased runoff in both natural and agricultural setting. Some forest areas of northern Iran are characterized by extreme water repellency, which makes their soil hydrological response highly complex. A field experiment was carried out to study the effect of water repellency on water movement and runoff in coarse-textured soil covered in Pinus teada in northwestern Iran. An adapted version of the popular agrohydrological model SWAP (Soil, Water, Atmosphere and Plant) has been used to evaluate the modelling of soil water flow and runoff generation. Water repellency was accounted for by multiplying the soil water content and the unsaturated hydraulic conductivity of soil with the F factor. This factor is equal to the volumetric fraction of soil occupied by preferential flow paths. Twenty-seven rainfall experiments were carried out under artificial rainfall using soils with various wettability levels and various rainfall intensities (14 mm/h, 27 mm/h and 34 mm/h). The result show that during the summer and early autumn (May to October) the soil became extreme repellent and runoff was formed when the soils were dry (θ<2%) and the rain intensity was more than 27 mm/h. Also, we investigated the role of the repellent layer on evaporation losses. The results of laboratory and simulation data confirm that water repellency reduces evaporation from the soil surface. Comparing modelling results with field data shows that an assumption of uniform flow in water repellent soil leads to an overestimation of soil water content and an underestimation of runoff. After incorporating water repellency in SWAP with the mobile–immobile concept, the agreement between measured and model simulation data considerably improved.