Soil water repellency as a potential parameter in rainfall-runoff modelling: experimental evidence at point to catchment scales from Portugal

Soil water repellency is generally thought to enhance runoff responses, thus representing a potentially important factor in hydrological modelling. Attempts to quantify its impacts have, however, either focused on soil profiles or plot scales and/or have been unable unequivocally to isolate repellency effects from other hydrologically important parameters. This research gap is addressed here by comparing responses of the same soil or terrain at a range of scales in highly repellent and wettable states, thereby limiting the impact of other variables. Hydrological responses of forest soils in north-central Portugal are assessed using laboratory wetting and rainfall simulation experiments at the point (0.002 m(2)) and microplot scales (0.12 m(2)), and field runoff responses from plot (16 m(2)) and catchment scales (0.33 km(2)). At the point scale, water repellency reduced soil wetting by a factor of >700 and increased mean runoff for a microplot from 1.5 to 53% during the first 9 min of a simulated rainstorm. At the field-plot scale impacts are evident, but less dramatic, with 16 of 45 of storms following dry, repellent conditions producing >3% overland flow compared with only one of 60 storms following moist, wettable conditions. In contrast, at the catchment scale, storm runoff responses were lower following dry, repellent, conditions than with moist, wettable conditions, although the time-to-peak was on average 62% shorter for the former compared with the latter conditions. The reduced correspondence between the scales is attributed mainly to increased capture of locally generated overland flow at the catchment scale by bypass routeways that are under-represented at point and plot scales. It is concluded that addressing water repellency in distributed hydrological modelling has considerable potential for improving predictions for affected regions, but that impacts may vary considerably from those found in the Portuguese case example. Key parameters for determining its impacts are its severity, spatial distribution, transient characteristics, and the character and spatial distribution of any bypass routeways and wettable subsoil. Copyright (C) 2003 John Wiley Sons, Ltd.