 |
| Titel |
Uncertainty of long-term CO2 flux estimates due to the choice of the spectral correction method |
| VerfasserIn |
Andreas Ibrom, Simon Geißler, Kim Pilegaard |
| Konferenz |
EGU General Assembly 2010
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 12 (2010) |
| Datensatznummer |
250041623
|
|
|
|
|
|
| Zusammenfassung |
| The eddy covariance system at the Danish beech forest long-term flux observation site at Sorø
has been intensively examined. Here we investigate which systematic and non-systematic
effects the choice of the spectral correction method has on long-term net CO2 flux estimates
and their components.
Ibrom et al. (2007) gave an overview over different ways to correct for low-pass filtering
of the atmospheric turbulent signal by a closed path eddy covariance system. They
used degraded temperature time series for spectral correction of low-pass filtered
signals. In this new study, correction for high-pass filtering was also included, which
made it anyway necessary to use model co-spectra. We compared different ways of
adapting different kinds of model co-spectra to the wealth of 14 years high frequency
raw data. As the trees grew, the distance between the sonic anemometer and the
displacement height decreased over time. The study enabled us to compare the
two approaches and different variants of them to give recommendations on their
use.
The analysis showed that model spectra should not be derived from co-spectra between
the vertical wind speed (w) and the scalars measured with the closed path system, i.e. CO2
and H20 concentrations, but instead with sonic temperature (T) w cospectra, to avoid
low-pass filtering effects on the estimation of the co-spectral peak frequency (fx). This
concern was already expressed earlier in the above mentioned study, but here we show the
quantitative effects.
The wT co-spectra did not show any height effect on fx as it was suggested in generally
used parameterizations. A possible reason for this difference is that measurements, like in all
forest flux sites, took place in the roughness sub-layer and not in the inertial sub-layer. At the
same time the shape of the relationship between fx and the stability parameter ζ
differed much from that of often used parameterizations (e.g. from Horst, 1997). The
shift of fx towards higher frequencies at stable atmospheric stratification was less
pronounced, resulting in less amount of correction for low-pass filtering at night
time and consequently higher annual net CO2 uptake estimates. On the other hand
our data indicate that the increase of fx already starts earlier than expected, i.e.
during the transition from unstable to neutral stratification. We derived an empirical
model of the shape of fx(ζ) with 4 parameters that is able to represent the observed
effects.
References
Horst, T.W. 1997. A simple formula for attenuation of eddy fluxes measured with
first-order-
response scalar sensors. Boundary-Layer Meteorology. 82:219-233.
Ibrom, A., E. Dellwik, N.O. Jensen, H. Flyvbjerg and K. Pilegaard 2007. Strong low-pass
filtering effects on water vapour flux measurements with closed-path eddy correlation
systems. Agricultural and Forest Meteorology. 147:140 -156.
Acknowledgements
This work has been funded by the EU-Infra structure project IMECC and the European
Erasmus Mundus program. We especially thank the Risø – DTU’s Wind Energy Department,
namely Ebba Dellwik and Søren W. Lund, for their strong technical and scientific
cooperation. |
|
|
|
|
|
|