 |
| Titel |
A statistical bias correction for climate model data: parameter sensitivity analysis. |
| VerfasserIn |
C. Piani, E. Coppola, L. Mariotti, J. Haerter, S. Hagemann |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250026167
|
|
|
|
|
|
| Zusammenfassung |
| Water management adaptation strategies depend crucially on high quality
projections of the hydrological cycle in view of anthropogenic climate change.
The goodness of hydrological cycle projections depends, in turn, on the
successful coupling of hydrological models to global (GCMs) or regional climate
models (RCMs). It is well known within the climate modelling community that
hydrological forcing output from climate models, in particular precipitation,
is partially affected by large bias. The bias affects all aspects of the
statistics, that is the mean, standard deviation (variability), skewness
(drizzle versus intense events, dry days) etc. The state-of-the-art approach to
bias correction is based on histogram equalization techniques. Such techniques
intrinsically correct all moments of the statistical intensity distribution.
However these methods are applicable to hydrological projections to the extent
that the correction itself is robust, that is, defined by few parameters that
are well constrained by available data and constant in time. Here we present
details of the statistical bias correction methodology developed within the
European project \"Water and Global Change\" (WATCH). We will suggest different
versions of the method that allow it to be taylored to differently structured
biases from different RCMs. Crucially, application of the methodology also
allows for a sensitivity analysis of the correction parameters on other gridded
variables such as orography and land use.
Here we explore some of these sensitivities as well. |
|
|
|
|
|
|