An Actively Heated Fiber Optics (AHFO) method to estimate soil moisture
is tested and the analysis technique improved on. The measurements
were performed in a lysimeter uniformly packed with loam soil with
variable water content profiles. In the first meter of the soil profi
le, 30 m of fiber optic cable were installed in a 12 loops coil.
The metal sheath armoring the fiber cable was used as an electrical
resistance heater to generate a heat pulse, and the soil response
was monitored with a Distributed Temperature Sensing (DTS) system.
We study the cooling following three continuous heat pulses of 120
s at 36 W m(-1) by means of long-time approximation of radial heat
conduction. The soil volumetric water contents were then inferred
from the estimated thermal conductivities through a specifically
calibrated model relating thermal conductivity and volumetric water
content. To use the pre-asymptotic data we employed a time correction
that allowed the volumetric water content to be estimated with a
precision of 0.01-0.035 (m(3) m(-3)). A comparison of the AHFO measurements
with soil-moisture measurements obtained with calibrated capacitance-based
probes gave good agreement for wetter soils [discrepancy between
the two methods was less than 0.04 (m(3) m(-3))]. In the shallow
drier soils, the AHFO method underestimated the volumetric water
content due to the longertime required for the temperature increment
to become asymptotic in less thermally conductive media [discrepancy
between the two methods was larger than 0.1 (m(3) m(-3))]. The present
work suggests that future applications of the AHFO method should
include longer heat pulses, that longer heating and cooling events
are analyzed, and, temperature increments ideally be measured with
higher frequency.