 |
| Titel |
Problems with modelling crops in the water cycle |
| VerfasserIn |
C. Jackisch, L. Samaniego, A. Hildebrandt |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250029565
|
|
|
|
|
|
| Zusammenfassung |
| Vegetation plays a major role both in the atmospheric and terrestrial water cycle. Cropland
covers a large proportion of the global terrestrial area (i.e. 38% globally in 2005, FAOSTAT).
Accordingly, crop models have become increasingly prominent for studying the impact of
Global Change both on economic welfare and on the influence of vegetation on climate
and feedback mechanisms with hydrological processes. By doing so, it is implied
that crop models properly reflect the crop-soil interactions, soil water balance and
vertical exchange with the atmosphere - and that parameters are to some degree
portable.
In this research we used data from eight lysimeters located at Brandis (Saxony, Germany),
which have been planted with the crops of the surrounding farm to test the capability of the
crop model in SWAP. The lysimeters contain different natural soil cores, leading to
substantially different yield. This experiment gives the opportunity to test the crop model
with regard to yield and the water cycle - i.e. whether a model calibrated with one crop and
one climate can be moved to a different soil.
When using the default parameters for the respective environment, the model does
neither quantitatively nor qualitatively reproduce the dynamics of the yield and
the water-balance for the different lysimeters. A joint calibration of plant and soil
parameters performed much better than that of a sequential calibration of the same
parameters. Moreover, the cross validation of parameter sets for the same crop
to a different soil indicates strong interrelations between the two parameter sets.
This suggests that the model is not portable, and needs to be calibrated at each
location. As we aim on dynamic crop modelling at the mesoscale with regionalized
parameters, we propose to work on understanding the orgin and effect of these
discrepancies. |
|
|
|
|
|
|