On the Relationship Between Soil Moisture, Soil Temperature, and Land Cover-Central Crops Research Station, Clayton, North Carolina, USA

A better understanding of soil moisture and temperature behavior among various categories of land cover is vital to maintaining both accurate archival of soil temperature and moisture data and further investigating the interactivity between pedospheric and atmospheric processes. Expected values for soil moisture and temperature are often contingent upon land cover type, namely vegetation-rich land, bare ground, rocky soil, etc. For instance, one would expect different rates of soil moisture loss between a plot of bare soil with no rooted vegetation as compared to a vegetation-rich plot with numerous, evaporation-inhibiting roots. A field experiment was conducted at a research farm located in central North Carolina, USA, where a coarse, loamy sand is the predominant soil characteristic. Two, one square meter plots of land were marked for this experiment. Three sensors were installed in each plot: a multilevel soil temperature sensor designed in-house at the State Climate Office of North Carolina, a CS-109 soil temperature probe, and a Delta-T ML3 soil moisture probe. The multilevel soil temperature probe has sensors at 10cm, 20cm, 30cm, and 40cm below the surface, while the CS-109 and Delta-T ML3 probes were installed at 20cm below the surface. All vegetation was cleared off of one plot, leaving only bare, loamy sand exposed at the surface, while the other plot was left to naturally redevelop. Soil moisture and temperature data were measured at a one-minute resolution, resulting in a detailed and uninterrupted time series spanning from June 2017 through September 2017. After initial measurements within the covered soil plot exhibited a distinct step-ladder drying trend, an additional Delta-T ML3 probe was installed at 35cm below the surface to determine if a layer of compacted soil due to long-term agricultural tillage practices was present. Data from this additional sensor would illustrate how soil moisture changes at a lower depth below this potential compacted soil layer. If a compacted layer of soil is present in the covered soil plot, the step-ladder trend could be caused by a cyclic process of rapid water uptake by roots during the daytime and subsequent pooling of water around the 20cm Delta-T ML3 probe. Analysis of the data shows that the covered land plot exhibits a well-defined, diurnal evaporation cycle that is not present with the bare soil plot. While the range between maximum and minimum soil moisture values were the same for both soil plots (0.41 m³m^-3), mean soil moisture in the covered plot was observed to be on average 0.015 m³m^-3, or 7.1% greater than the moisture in the bare soil. Response curves for the soil moisture in both plots responded to precipitation as expected, with bare soil absorbing water quicker and drying out at a faster rate than that of the covered soil. Soil moisture values in the covered plot had a variance greater than twice that of the bare soil, likely due to the step-ladder dehydration process observed most strongly during peak daytime solar radiation. The soil moisture trend from the additional Delta-T ML3 probe showed a smooth dehydration response curve congruent to the soil moisture trend observed within the bare soil plot. This insight into how soil moisture is changing at a 35cm depth supports the likelihood of a compacted layer of soil causing perturbations in the covered soil moisture dehydration around 20cm. Soil temperature measurements responded as expected with soil in the bare plot exhibiting a greater overall temperature range and higher mean temperature than the soil in the covered plot. The range between maximum and minimum temperature observed in the bare soil plot was 22.48℃ compared to only 18.01℃ in the covered soil plot. Bare soil temperature was consistently warmer during the daytime hours and cooler during nighttime hours, averaging 0.64℃ warmer than the covered soil throughout the experiment. Given that soil moisture sensors are installed at a 20cm standard at various research farms across North Carolina, further analysis may support the need for installation of additional probes at varying depths to observe different rates and behaviors of soil moisture loss. Multilevel soil temperature sensors are already commonplace and are often collocated with soil moisture probes, but the scope of interdependence between land cover and soil temperature requires further analysis. With a multitude of soil texture combinations and vegetative root systems present in the pedosphere, multiple levels of soil moisture monitoring and further research into soil temperature behavior would yield a better understanding of the relationship between land cover, soil temperature, and soil moisture.