Comparison of techniques for measuring the water content of soil and other porous media

Abstract

The measurement of water in soil on a potential, gravimetric or volumetric basis is considered, with studies concentrating on the measurement of water by dielectric and neutron moderation methods. The ability of the time-domain reflectometry technique to measure water content simultaneously at different spatial locations is an important advantage of the technique. The reported apparent dielectric by the TRASE� time-domain reflectometer and Pyelab time-domain reflectometry systems is sensitive to change in extension cable length. In some soil, e.g. a commercial sand, the response to increasing extension length of extension cable is linear. For other soil a linear response occurs for certain lengths of cable at different moisture contents. A single model accounting for clay content, extension cable length, time-domain reflectometry system, probe type and inherent moisture conditions explained 62.2 % of variation from the control (0 m extension) cable. The extension cable causes a decrease in the returning electromagnetic-wave energy; leading to a decline in the slope used in automatic end-point determination. Calibration for each probe installation when the soil is saturated, and at small water contents is recommended. The ability of time-domain reflectometry, frequency-domain and neutron moderation techniques in measuring soil water content in a Brown Chromosol is examined. An in situ calibration, across a limited range of water contents, for the neutron moderation method is more sensitive to changing soil water content than the factory supplied 'universal' calibration. Comparison of the EnviroSCAN� frequency-domain system and the NMM count ratio indicates the frequency-domain technique is more sensitive to change in soil water conditions. The EnviroSCAN� system is well suited to continuous profile-based measurement of soil water content. Results with the time-domain reflectometry technique were disappointing, indicating the limited applicability of time-domain reflectometry in profile based soil water content measurement in heavy-textured soil, or soil with a large electrical conductivity. The method of auguring to a known depth and placement of the time-domain reflectometry probe into undisturbed soil is not recommended. A time-domain reflectometry system is adapted for in situ measurement of water in an iron ore stockpile. The laboratory calibration for water content of the processed iron ore compares favourably to a field calibration. In the field study, the 28 m extension cable used to connect the probes to the time-domain reflectometry affected the end-point determination of the time-domain reflectometry system. To account for this, 0.197 should be subtracted from the reported apparent dielectric before calculation of volumetric moisture content.