Evaluation of Capacitance and TDR Soil Moisture Sensors in Undisturbed Soils Across the State of Mississippi.

Real-time monitoring of water content, temperature, and salinity at different depth in soil profiles is very important for irrigation scheduling and crop management as well as modeling of crop growth, hydrology, weather/climate impacts and hazard analyses. Soil moisture sensors are widely used in agriculture, hydrology, ecology, geophysics, and other areas. Sensors and probes, based on capacitance and time domain transmission, are the most commonly used sensors to monitor soil water, temperature and electrical conductivity.  Capacitance sensors CS655 (Campbell Sci., Inc) and GS1 (Decagon Devices, Inc), and time domain reflectometry (TDR315, Acclima, Inc) are three of the most robust sensors with stainless steel needles which are durable for installation and measurement in fields. Therefore, the three soil moisture sensors were assessed in six undisturbed soils across the state of Mississippi in the United States. In the laboratory, the three types of sensors were evaluated in three types of undisturbed soils from west MS Delta (Tunica clay, Commerce sandy loam, and Sharkey clay loam) and three from the east MS Blackland Prairie (Brooksville silty clay, Leeper sandy loam and Okolona loam). A PVC tube with sharp edge was hammered down 15 cm from top soils. After those undisturbed soils were saturated in a lab, CS655, TDR315 and GS1 were vertically inserted into those saturated soils and wired to CS1000 dataloggers, then placed on a scale with 0.1 g resolution. Scale readings were periodically recorded. As readings were relatively constant, the soils in tubes were dried for 48 hrs at 105ºC in an oven. Those determined soil volumetric water contents (Qv) were used as standards to compare with values measured by each sensor. The three types of probes were also assessed in top 15 cm and 15-30 cm layer of Brooksville silty clay soil in a field two times, during soybean growing season and after harvest, at the experiment station of Mississippi State University in Noxubee County, MS. Soil samples were periodically taken in the two layers and oven dried at 105 ºC for soil gravimetric water content (Qm). Three soil core samples were taken in each plot to measure bulk density for accurately convert Qm to Qv for comparisons with Qv measured by sensors. Results revealed that CS655, TDR315 and GS1 over estimated Qv in all six soils in a laboratory, field assessment for Brooksville silty clay showed that TDR315 under estimated Qv. Qv measured by CS655 were higher than TDR315 and GS1. It is not a simple linear relationship between Qv measured by sensors and gravimetric method.