 |
| Titel |
Value of eddy-covariance data for individual-based, forest gap models |
| VerfasserIn |
Edna Roedig, Matthias Cuntz, Andreas Huth |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250096788
|
| Publikation (Nr.) |
 EGU/EGU2014-12307.pdf |
|
|
|
|
|
| Zusammenfassung |
| Individual-based forest gap models simulate tree growth and carbon fluxes on large time
scales. They are a well established tool to predict forest dynamics and successions. However,
the effect of climatic variables on processes of such individual-based models is uncertain (e.g.
the effect of temperature or soil moisture on the gross primary production (GPP)).
Commonly, functional relationships and parameter values that describe the effect
of climate variables on the model processes are gathered from various vegetation
models of different spatial scales. Though, their accuracies and parameter values
have not been validated for the specific model scales of individual-based forest gap
models.
In this study, we address this uncertainty by linking Eddy-covariance (EC) data and a forest
gap model. The forest gap model FORMIND is applied on the Norwegian spruce
monoculture forest at Wetzstein in Thuringia, Germany for the years 2003-2008. The original
parameterizations of climatic functions are adapted according to the EC-data. The time step
of the model is reduced to one day in order to adapt to the high resolution EC-data. The
FORMIND model uses functional relationships on an individual level, whereas the
EC-method measures eco-physiological responses at the ecosystem level. However, we
assume that in homogeneous stands as in our study, functional relationships for both
methods are comparable. The model is then validated at the spruce forest Waldstein,
Germany.
Results show that the functional relationships used in the model, are similar to those observed
with the EC-method. The temperature reduction curve is well reflected in the EC-data, though
parameter values differ from the originally expected values. For example at the freezing
point, the observed GPP is 30% higher than predicted by the forest gap model. The response
of observed GPP to soil moisture shows that the permanent wilting point is 7 vol-% lower
than the value derived from the literature. The light response curve, integrated over the
canopy and the forest stand, is underestimated compared to the measured data. The
EC-method measures a yearly carbon balance of 13mol(CO2)m-2 for the Wetzstein site.
The model with the original parameterization overestimates the yearly carbon balance
by nearly 5mol(CO2)m-2 while the model with an EC-based parameterization
fits the measured data very well. The parameter values derived from EC-data are
applied on the spruce forest Waldstein and clearly improve estimates of the carbon
balance. |
|
|
|
|
|
|