QUANTIFYING ERRORS IN HEAT FLUX MEASUREMENTS USING A VERTICAL SEQUENCE OF THERMOCOUPLE SENSORS

Estimating the heat flux at the land surface in geothermal areas can be a difficult

task. In part, this is due to the fact that the thermal conductivity of the soils is

spatially variable. More importantly, the time-varying atmospheric boundary condition

propagates a quasi-periodic signal into the subsurface, causing the near surface flux to

vary temporally such that it generally reverses on a diurnal time-scale. Under these

conditions, the most accurate way to estimate the annual average heat flux is to install a

vertical sequence of thermocouple sensors in the shallow subsurface, usually consisting

of four to six sensors distributed over about 1 m of depth. With accurate data on the

subsurface temperature profile and the soil thermal conductivity, it is possible to correct

for the time-varying surface boundary condition. However, it is common practice to

attach the thermocouple sensors to a metal rod, which is then driven into the soils,

and the perturbation introduced to the near-surface temperature profile has not been

previously evaluated. Here, we evaluate the impact of introducing a high-conductivity

element on soil heat-transfer, and offer guidelines for collecting shallow temperature

profile time-series data.