Using neural networks for calibration of time-domain reflectometry measurements

Time-domain reflectometry (TDR) is an electromagnetic technique for measure ments of water and solute transport in soils. The relationship between the TDR-measured dielectric constant (K-a) and bulk soil electrical conductivit y (sigma (a)) to water content (theta (w)) and solute concentration is diff icult to describe physically due to the complex dielectric response of wet soil. This has led to the development of mostly empirical calibration model s. In the present study, artificial neural networks (ANNs) are utilized for calculations of theta (w) and soil solution electrical conductivity (sigma (w)) from TDR-measured K-a and sigma (a) in sand. The ANN model performanc e is compared to other existing models. The results show that the ANN perfo rms consistently better than all other models, suggesting the suitability o f ANNs for accurate TDR calibrations.