Feasible calibration methods are required for the application of commercial sensors to field monitoring of soil water content and solution concentration. Simple in situ calibration methods for the 5TE sensor are proposed for the estimation of soil water content (θ) and 1 : 5 soil to water extract electrical conductivity (EC1:5). θ can be obtained from 5TE sensor readings with a correction fac-tor C, which is available from several soil samples taken during monitoring. EC1:5 can be obtained from a linear function of soil pore water electrical conductivity (ECp). ECp is derived from 5TE sensor readings using the Hil-horst equation, and the gradient and intercept of the linear function are also calculable from a small number of soil samples taken during monitoring. The proposed calibra-tion methods were applied to monitoring of θ and EC1:5 in a tomato field under drip fertigation, and the efficacy of the methods was confirmed by comparison with methods using conventional lab-based calibration.
When the upper part of a soil solution has a larger density than the lower part, an unstable layer devel-ops, and convection may occur. However, our theoretical understanding of this convective flow pattern is limited, given the complexity of its governing equations. In this study, we conducted a simple experiment to generate con-vection, and used linear stability theory to explain its flow pattern. Glass beads with 0.2-mm diameter were placed in a 6×14×1 cm rectangular container. The lower half of this container was saturated with water, while the upper half
was saturated with a concentrated and regulated NaCl so-lution. The resulting solution density interface undulated, and finger convection emerged. We found that the average speed of the fingers could be expressed using the equa-tion: hydraulic conductivity×solution density difference / solution density of the lower part. This is consistent with predictions from linear stability theory. We also found that the number of fingers had a non-linear relationship with the solution density difference. The number of fingers corre-sponded to the wave number for the larger steady solution, when the length of the density gradient area was twice the distance from the density interface to the front of the fin-ger.