Modelling soil evaporation in an agroforestry system in Kenya

Soil evaporation measurements from bare soil and shaded soil under an agroforestry tree canopy were used to construct a model to predict soil evaporation with and without tree shade. It was found that a simple daily time step model based on the Ritchie (1972)approach was unable to predict daily soil evaporation accurately, but was capable of providing good estimates of cumulative soil evaporation over hydrologically significant periods (weeks–months). This model was used to show how trees could reduce annual soil evaporation directly beneath their canopy by an average of 35% (compared to completely bare soil), equivalent to 21% of rainfall. In sparse agroforestry tree canopies the area average saving is smaller, depending on tree leaf area index (LAI). The model also demonstrated how annual saving in soil evaporation due to a tree canopy might vary with rainfall, with a maximum of around 180 mm being achieved once rainfall exceeded ∼1000 mm year −1. This saving in soil water is very significant and will help offset the enhanced evaporative losses associated with tree canopies due to interception and re-evaporation of rainfall or as tree transpiration.