Search

A number of automated techniques for point measurement of soil moisture content have been developed to an operational level over the past few decades. While each of those techniques have been individually evaluated by the thermogravimetric (oven drying and weighing) method, typically under laboratory conditions, there have been few studies which have made a direct comparison between the various techniques, particularly under field conditions. This paper makes an inter-comparison of the Virrib®, Campbell Scientific CS615 reflectometer, Soil Moisture Equipment Corporation TRASE® buriable- and connector-type time domain reflectometry (TDR) soil moisture sensors, and a comparison of the connector-type TDR sensor with thermogravimetric measurements for data collected during a 2-year field study. Both qualitative and quantitative comparisons between the techniques are made, and comparisons made with results from a simple water balance ‘bucket’ model and a Richards equation based model. It was found that the connector-type TDR sensors produced soil moisture measurements within the ±2.5% v/v accuracy specification of the manufacturer as compared to thermogravimetric data when using the manufacturer''s calibration relationship. However, comparisons with the water balance model showed that Virrib and buriable-type TDR sensors yielded soil moisture changes that exceeded rainfall amounts during infiltration events. It was also found that the CS615 reflectometer yielded physically impossible soil moisture measurements (greater than the soil porosity) during periods of saturation. Moreover, the buriable-type TDR measurements of soil moisture content were systematically less than the Virrib measurements by approximately 10% v/v. In addition to the good agreement with thermogravimetric measurements, the connector-type TDR soil moisture measurements yielded the best agreement with Richards equation based model predictions of soil moisture content, with Virrib sensors yielding a poor agreement in the deeper layers. This study suggests that connector-type TDR sensors give the most accurate measurements of soil moisture content out of the sensor types tested. [Copyright &y& Elsevier]

Summary: It is well documented that capacitance-based soil moisture sensor measurements are particularly influenced by particle size distribution, density, salinity, and temperature of a soil, in addition to its moisture content. Moreover, the equations provided by manufacturers of soil moisture sensors are often only applicable to a limited number of soil types, thus yielding significant errors when compared with gravimetric measurements for observations in real soils. This limitation makes site-specific calibrations of such sensors necessary. Consequently, development of a general equation provides the possibility to derive the needed parameters from information such as soil type or particle size distribution. This paper describes the development of a general equation for the Campbell Scientific CS616 Water Content Reflectometers using data from sensors installed throughout the Goulburn River experimental catchment. It is subsequently tested using monitoring sites in the Murrumbidgee Soil Moisture Monitoring Network, which were not part of the original development; both monitoring networks are located in south-eastern Australia. Previously developed equations for temperature correction and soil moisture estimation using the Campbell Scientific CS615 Water Content Reflectometer are adapted to the new CS616 sensor. Moreover, relationships between readily available soil properties and the parameters of the general equations are derived. It is shown that the general equations developed here can be applied to data collected in the field using only information on the soil particle size distribution with an RMSE of around 6%m3/m3 (<1%m3/m3 under laboratory conditions; which is a significant improvement in comparison to 14%m3/m3 when using the manufacturer’s equations). [Copyright &y& Elsevier]