WP1200/WP1300 Diagnostic analysis of
ensemble integrations performed under the PRISM (Predictability experiments
for the Asian summer monsoon) project. (in collaboration with CINECA,
ECMWF, UREDMY).
The dominant patterns of
interannual variability of the Asian summer monsoon circulation (as simulated
in the PRISM experiments) have been determined by means of a SVD analysis
of 850-hPa wind and rainfall in the region (60-120E, 40N-20S). The two
leading SVD pattern resemble some of the dominant variability patterns
found in observational studies based on re-analyses (Annamalai et al,
MWR 1999; Sperber et al, QJRMS 2000). However, the modelled distribution
of variance is quite different from the observed distribution, with a
very large fraction (especially for the wind field) accounted for by the
first mode. This mode (which is associated with a meridional shift of
the TCZ in the Indian Ocean) shows a relatively low degree of reproducibility
in ensembles with observed SST, and the agreement between ensemble-mean
and re-analysis values of the correspondent time-coefficient is poor.
Conversely, the second mode (quite similar to the observed composite in
years of anomalous All-India Rainfall) is strongly influenced by the ENSO
cycle, and a good agreement is found between the observed and ensemble-mean
coefficient. An analysis of the relationship between these interannual
variability modes and the intraseasonal variability of modelled rainfall
(as deduced from 5-day mean field) is under way.
WP5000 Collaboration with research
institutions in monsoon-affected countries.
Scientists from three institutions
interested in research collaborations under PROMISE (from Brazil, India
and the Philippines)ã have been
nominated Associate Members of ICTP in 2000. This entitles them to a number
of funded visits (each lasting about 2-3 months) to ICTP in the next six
years.
Another ICTP Associate
Member (S. Mwandoto, from the Kenia Meteorological Department in Nairobi)
has visited ICTP from March to July 2000, working on regional verifications
of rainfall simulated by the PROVOST ensemble simulation. His visit has
provided a useful 'test-case' for the development of simple validation
and training tools for scientists from developing countries.
Specifically, a preprocessing
of rainfall data from ERA and the ECMWF PROVOST ensembles (as included
in the ECMWF CD-ROMs) has been performed, in order to allow an easy comparison
between rainfall anomalies from ERA and the ensemble means
in different years/seasons. From this dataset, a GRADS script produces
colour plots of the two fields for specified periods (year, season, range
of 10-day means) and regions (five longitude bands can be selected). It
is planned to develop a web-based interface to simple verification tools
like this one, in collaboration with CINECA, (to be accessed from the
PROMISE archive web page), and to apply them to other experimental datasets
(e.g. PRISM).
Finally, a Summer Colloquium
on "Land-Atmosphere Interactions and the Hydrological Cycle",
including a training workshop and a research conference, has been planned
for 28 May - 15 June 2001. This event (with J. Slingo from UREADMY as
Co-Director) will address a number of PROMISE-related issues, and will
act as the main PROMISE meeting of 2001. |