Soil moisture, temperature and heat flux measurements using Distributed Temperature Sensing

Distributed Temperature Sensing is a powerful, relatively new tool in environmental monitoring. Fibre-optic cables are used as thermal sensors, yielding accurate temperature observations every meter or so, along cables that can be several kilometers long. They have been employed in a huge number of applications including field investigations of streamflow, snow dynamics, and landslides. We have focused on their potential use to observe states of interest in land-atmosphere interactions. Results will be shown from two experiments. In the first, temperature measurements were made at a test site in the Netherlands. Because soil thermal properties depend on soil moisture, an inversion approach was first used to estimate soil moisture in the top 10cm of the soil. Changes in thermal diffusivity/conductivity were observed in response to precipitation. Due to non-uniqueness in the relationship between diffusivity and soil moisture, it proved difficult to infer a single value of soil moisture from the thermal diffusivity. A sensitivity study demonstrated that the results were sensitive to the cable depths and the certainty therein. Data assimilation was explored as an alternative to the inversion method. A dual state-parameter estimation approach, based on the Ensemble Kalman filter, was used to assimilate the temperature measured at two depths into a coupled heat-moisture model. This allowed us to estimate the temperature, soil moisture and soil heat flux profile at the resolution of the model to a depth of 50cm. In the second experiment, cables were plowed to three depths at an experimental site in Reno, Nevada and the temperature measurements were used to investigate the horizontal spatial variability of ground heat flux. Surface heat flux was shown to be highly variable, in some cases up to 100% within a few meters.