Modeling infiltration into water repellent soil

Abstract

Infiltration is an important hydrological process impacting ecology, forestry, agronomy, civil- and environmental engineering. Most infiltration models assume soils to be “wettable”, i.e., water in the soil forming an effective contact angle with the soil matrix that is close to zero. For a range of applications, e.g., infiltration into organic-rich soils or soils that turned water repellent due to fire, the “wettability assumption” no longer holds. Hence, the need for an infiltration model that can take soil water repellency into account. This study proposes a process-based approach for modeling infiltration into water repellent soil using the concepts of effective contact angle and sorptivity. The approach was developed using the Green-Ampt infiltration model but can be easily adapted for other process-based infiltration models such as Philip or Smith-Parlange. The infiltration model demonstrates the considerable impact of soil water repellency on infiltration, also for subcritically-water repellent soils, i.e., soils with effective contact angles <90°. It illustrates the non-linear relationship between infiltration rate and effective contact angle with effective contact angles >70° having a much larger impact on infiltration rate than effective contact angles <70°. The model also indicates that due to gravity, infiltration could occur into super-critically water repellent soil (i.e., soil with effective contact angles ≥90°), even with zero hydraulic head at the soil surface. Infiltration at zero hydraulic head, however, likely ceases at effective contact angles between 91° and 101°, depending on the amount of cumulative infiltration. All infiltration simulations showed decreasing infiltration rates with increasing soil water repellency expressed as effective contact angle or sorptivity at any level of cumulative infiltration. Finally, the water repellency effects on infiltration rates for short-duration, high intensity storms—a critical situation commonly associated with wildfire and flooding—were illustrated.