 |
| Titel |
Using Distributed Temperature Sensing for evaporation measurements:
background, verification, and future applications. |
| VerfasserIn |
Bart Schilperoort, Miriam Coenders-Gerrits, Tara van Iersel, Cesar Jiménez Rodríguez, Willem Luxemburg, Cesar Cisneros Vaca, Murat Ucer |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250146549
|
| Publikation (Nr.) |
 EGU/EGU2017-10577.pdf |
|
|
|
|
|
| Zusammenfassung |
| Distributed temperature sensing (DTS) is a relatively new method for measuring latent and
sensible heat fluxes. The method has been successfully tested before on multiple sites
(Euser, 2014). It uses a glass fibre optic cable of which the temperature can be
measured every 12.5cm. By placing the cable vertically along a structure, the air
temperature profile can be measured. If the cable is wrapped with cloth and kept wet (akin
to a psychrometer), a vertical wet-bulb temperature gradient over height can be
calculated. From these dry and wet-bulb temperatures over the height the Bowen ratio is
determined and together with the energy balance the latent and sensible heat can be
determined.
To verify the measurements of the DTS based Bowen ratio method (BR-DTS) we
assessed in detail; the accuracy of the air temperature and wet-bulb temperature
measurements, the influence of solar radiation and wind on these temperatures, and a
comparison to standard methods of evaporation measurement. We tested the performance of
the BR-DTS on a 45m high tower in a tall mixed forest in the centre of the Netherlands in
August. The average tree height is 30m, hence we measure temperature gradients above, in,
and underneath the canopy.
We found that solar radiation has a significant effect on the temperature measurements
due to heating of the cable coating and leads to deviations up to 2∘
C. By using cables with
different coating thickness we could theoretically correct for this effect, but this introduces
too much uncertainty for calculating the temperature gradient. By installing screens the effect
of direct sunlight on the cable is sufficiently reduced, and the correlation of the cable
temperature with reference air temperature sensors is very high (R2=0.988 to 0.998). Wind
speed seems to have a minimal effect on the measured wet-bulb temperature, both below and
above the canopy.
The latent heat fluxes of the BR-DTS were compared to an eddy covariance
system using data from 10 days, with quality control applied to both methods. When
comparing the daytime values, there is a high correlation (R2=0.75), a low bias (mean
difference of ±15W/m2) and a good accuracy (standard deviation of the difference
of 40W/m2) for both the latent and sensible heat flux. This can lead to a small
error. Nonetheless, the results show that when the system is set up with care, and by
eliminating sources of errors, the DTS based Bowen ratio is in agreement with an eddy
covariance system, even above a tall forest canopy, which is notoriously hard to
measure.
Further applications of the DTS data in evaporation measurement studies are the
flux-variance method (where the standard deviations of the air temperature and absolute
humidity are used to estimate the sensible and latent heat fluxes), the surface-renewal method,
and correcting the Bowen ratio for the non-unity of the eddy diffusivity ratios. These can
all be used to gather additional data on the evaporation to increase the accuracy. |
|
|
|
|
|
|