Evaluation of a Heat Dissipation Sensor for In Situ Measurement of Soil Temperature

Scientists at the Oklahoma Mesonet analyzed the Campbell Scientific (CS) 229 heat dissipation matric potential sensor to determine if it could be used as a dual‐purpose sensor to measure soil temperature in addition to soil matric potential. The objectives of the study were to identify how the installation wiring of the CS 229 impacts the accuracy of temperature measurements, to understand how the CS 229 temperature measurements compared with existing soil temperature measurements taken by BetaTHERM thermistors, and to determine the spatial variabilities of soil temperature measurements in situ. An improved wiring technique for the CS 229 sensor greatly reduced the mean squared error of temperature measurements from 0.20 to 0.01°C. The BetaTHERM and CS 229 sensors yielded strong correlations (RMSD of 0.96°C at 5 cm) and linearities (R2 of 0.99 at 5 cm) with a small bias (mean difference of 0.19°C at 5 cm toward the CS 229) in the field. The two sensor types also had low variability in their differences (average RMSE of 0.06°C for BetaTHERM and 0.09°C for CS 229) when analyzed in a controlled laboratory environment. The variability (e.g., standard deviation) of the daily average 5‐cm soil temperatures measured within a 10‐m transect was 0.29°C, while less variability was observed at the 10‐ and 30‐cm depths. The analyses showed that the CS 229 would provide soil temperature measurements of comparable accuracy to that of existing BetaTHERM sensors. The overall results of the experiments demonstrated that the CS 229 is a viable soil temperature sensor when properly calibrated and installed.