![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
Sensitivity analysis of a soil methane emission model. |
VerfasserIn |
J. van Huissteden, D. M. D. Hendriks, A. M. R. Petrescu |
Konferenz |
EGU General Assembly 2009
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 11 (2009) |
Datensatznummer |
250021774
|
|
|
|
Zusammenfassung |
CH4 emission from wetland soils depends on a large number of parameters, related to soil,
vegetation and microbial population characteristics. Several of theses parameters are difficult
to determine by measurement, or the available amount of data is limited. For estimation of
CH4 emission using models, model confirmation, parameter estimation and parameter
sensitivity should be assessed, in particular if these models are to be applied for
modelling emissions over a range of wetland and soil types. We present the result of
sensitivity analysis and parameter estimation of a model of wetland CH4 emission
(PEATLAND-VU). The validation data are derived from various wetland sites,
ranging from temperate to arctic, and with different management conditions. For
the analysis we used the GLUE (Generalized Likelihood Uncertainty Estimation)
approach.
The performance of the model showed large differences between the validation sites. For
one site, the model did not perform better than an emission estimate based on the mean of the
data, in other cases the model fit to the data was satisfactory. The site for which the
performance was poorest (model fit not better than weighed average of the data) was a site
with a deviating management history, being converted from agricultural land to wetland
recently; the other sites were more natural wetlands.
In general, the parameter sets of model simulations were highly non-unique; good model
fits could be obtained with parameter sets that were widely different. The model parameters
that proved to be highly identifiable in sensitivity analysis were vegetation-related
parameters, while the model showed a comparatively low sensitivity to soil parameters. This
suggests that for global scale modelling of CH4 fluxes vegetation parameters are most
important. |
|
|
|
|
|