 |
| Titel |
When will climate scenario runs show a statistically significant change in drought indices? |
| VerfasserIn |
P. Thejll, C. Fox Maule, J. Hesselbjerg Christensen |
| Konferenz |
EGU General Assembly 2012
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 14 (2012) |
| Datensatznummer |
250062036
|
|
|
|
|
|
| Zusammenfassung |
| Climate models can be used to evaluate the consequences of various
emission scenarios. A common application is to look at scenario runs
and find out what will happen to e.g. temperature or precipitation in
the future. Typically, the result is that some model observable will
increase or decrease in the mean as time passes in the model. Given
the model level of variability it is possible to ask when a particular
quantity rises (or falls) so much that the change, relative to some base
period, can be said to be statistically significant. We investigate when
the mean of two drought indices - the Standardized Precipitation Index
(SPI) and the self-calibrated Palmer Drought Severity Index (scPDSI) -
change by a statistically significant amount over regions of Europe in
the A1B scenario runs from the ENSEMBLES project. Assessment of the
statistical significance hinges on a thorough understanding of the
number of degrees of freedom - N - in the model fields - in time and
space. We use a simple approach and set upper and lower limits on N, and
calculate the probabilities corresponding to model field mean changes. The
results are that a statistically significant change in the mean scPDSI
index occurs in some of the European regions in the models - with good
uniformity across the range of models. For the SPI a similar result is
found, but less distinctly, with the most similar results in the NW and
SW tending to occur a little later than for the scPDSI. Typically the
clearest change-signals are seen in the North-West and South-West of
Europe. The scPDSI depends on temperature as well as precipitation so
the clarity of the signal in the scPDSI may be due to the more robustly
modelled future temperature changes. A similar analysis of the changes of
drought-index variance has not yielded any clear results. We use these
findings to discuss the extent to which scenario runs may be used in
predictions of real-world drought-condition changes across Europe. |
|
|
|
|
|
|