No abstract

No abstract

No abstract

Core Ideas A new thermo‐TDR sensor can determine soil thermal properties, water content, bulk density, porosity, and air‐filled porosity. The new theories are used to analyze the heat‐pulse data and TDR waveforms. The new sensor provides greater sensing volume and more accurate results than previous designs. The thermo‐time domain reflectometry (thermo‐TDR) technique is valuable for monitoring in situ soil water content (θ), thermal properties, bulk density (ρb), porosity (n), and air‐filled porosity (na) in the vadose zone. However, the previous thermo‐TDR sensor has several weaknesses, including limited precision of TDR waveforms due to the short probe length, small measurement volume, and thermal property estimation errors resulting from finite probe properties not accounted for by the heat pulse method. We have developed a new thermo‐TDR sensor design for monitoring θ, thermal properties, ρb, n, and na. The new sensor has a robust heater probe (outer diameter of 2.38 mm and length of 70 mm) and a 10‐mm spacing between the heater and sensing probes, which provides a sensing volume three times larger than that of the previous sensor. The identical cylindrical perfect conductors and the tangent line–second‐order bounded mean oscillation theories were applied to analyze the raw data. Laboratory tests showed that θ values determined with the new sensor had a RMSE of 0.014 m3 m−3 compared with 0.016 to 0.026 m3 m−3 with the previous sensor. Soil thermal property estimates with the new sensor agreed well with modeled values. Soil ρb, n, and na derived from θ and thermal properties were consistent with those derived from gravimetric measurements. Thus, the new thermo‐TDR sensor provides more accurate θ, thermal properties, ρb, n, and na values than the previous sensor.

Soil thermal diffusivity κ is an essential parameter for studying surface and subsurface heat transfer and temperature changes. It is well understood that κ mainly varies with soil texture, water content θ, and bulk density ρb, but few models are available to accurately quantify the relationship. In this study, an empirical model is developed for estimating κ from soil particle size distribution, ρb, and degree of water saturation Sr. The model parameters are determined by fitting the proposed equations to heat-pulse κ data for eight soils covering wide ranges of texture, ρb, and Sr. Independent evaluations with published κ data show that the new model describes the κ(Sr) relationship accurately, with root-mean-square errors less than 0.75 × 10−7 m2 s−1. The proposed κ(Sr) model also describes the responses of κ to ρb changes accurately in both laboratory and field conditions. The new model is also used successfully for predicting near-surface soil temperature dynamics using the harmonic method. The results suggest that this model provides useful estimates of κ from Sr, ρb, and soil texture.