Assessing soil water content variability through active heat distributed fiber optic temperature sensing

Abstract

Soil spatial variability is a key point for the sustainable water management in agriculture. Fractal techniques provide proper tools to analyze soil spatial variability searching for statistical self-similarity patterns among different scales. Although they have been extensively applied to study the soil properties variability, its applicability for the soil water content (SWC) distribution is complicated because requires many data difficult to obtain with the typical point soil water sensors. Recently, a fiber optic distributed temperature sensor has been used to measure soil thermal properties which relate to SWC. These sensors provide large amount of data with high spatial and temporal resolution, thus filling the gap of point soil water sensors. In the present work, soil temperature was measured with a Distributed Temperature Sensing (DTS) and SWC was estimated by different fitting functions which have been studied with focus on spatial variability. Temperature was measured in a 133 m fiber optic cable laid in a sandy soil field plot. A The Active Heat Fiber Optic (AHFO) method was used, with 12 cm sampling resolution, and heat pulses (19,4 W/m during 2 min) were applied. The temperature data were correlated to SWC, considering the integration of temperature during the heat pulse Tcum, and then the datasets Tcum-SWC were fitted to the best fit statistical function (exponential, potential and polynomial). The results showed that the Tcum distribution presented a non-Gaussian pattern. Additionally, highly anti-persistent patterns have been detected for the larger spatial scaling lags. The function`s performance was different thus, the exponential function reproduced better the absolute moments of the temperature profile but it failed reproducing the non-Gaussian behavior.