Evaluating Soil Water Content based on ANN and Time Domain Reflectometry

Abstract

Soil volumetric water content measurement always use Time Domain Reflectometry, which exploits the difference in dielectric constant values between the solid phase, air phase and liquid phase. We tried to use the empirical data and model to fit the Time Domain Reflectometry for different soils textures and used artificial neural network (ANN) to measure the parameters for ten different heavy texture soil types. The measures such as the Dielectric constant, bulk density, clay content, silt content, sand content and organic matter content were detected. A comparative study among ANN models and various existing empirical models was also carried out. In experiment, the ANN models gave better predictions than empirical models. The ANN model performed superior than both empirical and physical models.