An Active-Distributed Temperature Sensing method for measuring groundwater flow velocities into streambed sediments at high spatial resolution.
Abstract
Understanding and quantifying groundwater and surface water interactions are key elements for the management
of water quality and quantity but also for the preservation of groundwater dependent ecosystems and riparian
habitat. Localizing and quantifying these exchanges remains challenging since exchange processes vary both in
time and space leading to an extreme heterogeneity in the distribution of fluid exchanges. Given the interest in
characterizing hyporheic flows and quantifying interactions between groundwater and surface water, we developed
a new method for measuring groundwater flow velocities in streambed sediments with an unprecedented high
spatial resolution.
The experimental setup consists in heating an armored fiber-optic cable, previously deployed along the
streambed within the sediments of a first-order and intermittent stream. According to heat transport principle,
the temperature increase along the cable depends mainly on the sediments thermal properties and on
groundwater flow velocities that may dissipate heat through heat advection. Thus, the temperature reached
after a given time and at a given location should be a function of local groundwater flow velocities. Thermal
response measured along heated the fiber-optic cable are interpreted by assuming that the heated cable can
be modelled as a Moving Infinite Line Source (MILS), i.e. a thermal source, using the analytical solution
developed by Metzger et al in 2004. To validate the use of the analytical solution and to estimate the uncertainty
and limits associated to the method, a 2D numerical model has been developed. This model simulates steady
state fluid flow and transient heat transfer using the Conjugate Heat Transfer module of COMSOL Multiphysics ®.
After few hours of heating, the measured temperature at steady state is particularly variable along the section
with temperature increases from 16 to 36C. By applying the analytical model, we show that it is possible
to reproduce field measurements by varying flow velocity only. Therefore, this suggests that the thermal response
variability can simply be associated with local variations of groundwater fluxes. These field tests allowed
characterizing the groundwater flow velocities distribution into the streambed as a lognormal distribution ranging
between 8e-7 and 6e-5 m/s.