Development of heat-pulse sensors for measuring fluxes of water and solutes under the root zone

Abstract

The objectives defined for this study were to: (1) develop a heat-pulse sensor and a heat-transfer model for leaching measurement, and (2) conduct laboratory study of the sensor and the methodology to estimate leaching flux. In this study we investigated the feasibility for estimating leachate fluxes with a newly designed heat-pulse (HP) sensor, combining water flux density (WFD) with electrical conductivity (EC) measurements in the same sensor. Whereas previous studies used the conventional heat pulse sensor for these measurements, the focus here was to estimate WFD with a robust sensor, appropriate for field settings, having thick-walled large-diameter probes that would minimize their flexing during and after installation and reduce associated errors. The HP method for measuring WFD in one dimension is based on a three-rod arrangement, aligned in the direction of the flow (vertical for leaching). A heat pulse is released from a center rod and the temperature response is monitored with upstream (US) and downstream (DS) rods. Water moving through the soil caries heat with it, causing differences in temperature response at the US and DS locations. Appropriate theory (e.g., Ren et al., 2000) is then used to determine WFD from the differences in temperature response. In this study, we have constructed sensors with large probes and developed numerical and analytical solutions for approximating the measurement. One-dimensional flow experiments were conducted with WFD ranging between 50 and 700 cm per day. A numerical model was developed to mimic the measurements, and also served for the evaluation of the analytical solution. For estimation WFD, and analytical model was developed to approximate heat transfer in this setting. The analytical solution was based on the work of Knight et al. (2012) and Knight et al. (2016), which suggests that the finite properties of the rods can be captured to a large extent by assuming them to be cylindrical perfect conductors. We found that: (1) the sensor is sensitive for measuring WFD in the investigated range, (2) the numerical model well-represents the sensor measurement, and (2) the analytical approximation could be improved by accounting for water and heat flow divergence by the large rods.