Geostatistical Stochastic Simulation of Soil Water Content in a Forested Area of South Italy

Abstract

Understanding and managing the spatial variability of surface soil moisture is a critical issue in solving numerous problems. Stochastic simulation was applied for assessing the probability that soil water content does not exceed a critical threshold that might cause water stress to a forest of laricio pine in an experimental plot located in south Italy. Soil water measurements were made at 152 points with 45 cm long rods using a time domain reflectometer. A probabilistic approach was used to assess the level of topsoil water depletion. It is based on the use of the geostatistical technique of conditional simulation to yield a series of stochastic images of equally probable spatial distributions of the soil water contents across the site. The proposed approach allowed the recognition of explicit uncertainty in soil water assessment and the incorporation of the likelihood that soil dryness may become critical for plants in site characterisation.