Presentazione di PowerPoint

1
INSTITUTE OF ATMOSPHERIC SCIENCES AND CLIMATE
PA0024
OVERVIEW The availability of the 4-D VAR ECMWF
reanalysis for the MAP Special Observing Period,
allows us to perform a numerical assessment of
its impact on high resolution forecasts in
comparison with the operational 3-D VAR ECMWF
analysis. By means of the BOLAM (hydrostatic) and
MOLOCH (non-hydrostatic) limited area models
operating at different resolutions and using
ECMWF analysis and reanalysis as initial
conditions, we evaluated the quantitative
precipitation forecasts in three MAP Intensive
Observing Periods using statistical scores
applied to the precipitation fields.
Goals Our purpose is on the one hand to test the
sensitivity of the precipitation forecast versus
analysis type by contrasting results obtained
with OPER.99 and MAP-RA. On the other hand to
compare the results obtained with different
limited area models, BOLAM and MOLOCH.
Verification In order to compute the statistical
scores on the 24 hour forecast precipitation for
the three cases, data from the MAP dataset
(http//www.map.ethz.ch) have been employed.

• The models
• BOLAM hydrostatic, primitive equations limited
  area model in s coordinates (Buzzi et al.,
  2003a), developed at ISAC-CNR in Bologna.

• The case studies
• In IOP-2b and IOP-8 the general flow impinging
  upon the Alps was southerly or south-westerly and
  substantial amounts of orographic precipitation
  fell on the southern flank of the chain.
• IOP-2b
• slightly unstable atmospheric thermodynamic
  profile on the upstream side of the Alps.
• convective rain occurred, with abundant
  accumulated values, generally larger than 200
  mm/24 h, observed at many locations along the
  southern side of the Alps.
• IOP-8
• the pre-existing cold air trapping over Po
  Valley favoured the establishment of stably
  stratified conditions.
• event characterized by smaller precipitation
  values, exceeding 100 mm/24 hours only in few
  areas.
• IOP-15
• intense lee cyclone formation south of the Alps
  (Buzzi et al., 2003a), characterized by strong
  and damaging mesoscale phenomena.
• precipitation occurences on the south side of
  the Alps from the Lago Maggiore area to the
  Veneto-Friuli pre-Alpine area and also the
  northern Appennines.

• Low Resolution 0.2x0.2, 38 vertical levels
• High Resolution 0.06x0.06, 44 vertical levels

• MOLOCH non-hydrostatic model recently developed
  at ISAC-CNR (Buzzi et al., 2003b) that integrates
  the fully compressible set of equations

• Very high resolution 2.2x2.2 km and 50 levels

Experimental set-up The 4-D VAR ECMWF reanalysis
(MAP-RA Keil and Cardinali, 2003) and the 3-D
VAR ECMWF analysis (OPER.99) are used to perform
the numerical experiments focusing on the MAP-SOP
episodes corresponding to IOP-2b, IOP-8 and
IOP-15. The same strategy is used for the three
case studies. Initial and boundary conditions for
the low resolution BOLAM runs are supplied by
ECMWF 6-hourly analyses (either OPER.99 or
MAP-RA). The high resolution BOLAM simulations
are nested in the lower resoluion ones. MOLOCH
runs are nested into BOLAM high resolution
experiments, with lateral boundary conditions
updated every hour.
The evaluation method The evaluation is made
using statistical scores generally applied to
precipitation fields, introducing a smoothing
criterion based on model derived probability
estimates. Let POi denote the observed
precipitation at the ith observation point. Let
us introduce the probability (frequency) function
?i(s) that the simulated precipitation at the ith
station be greater than the threshold s. For a
perfect model, ?i(s)H(POI-s), H being the
Heaviside function. The number of hits, false
alarms and misses can be defined, respectively,
as follows

Smoothed scores

IOP-2b
IOP-2b
IOP-8
IOP-8

• Final remarks
• The effects of MAP-RA are not easily
  interpreted. However, in the case of IOP-15, the
  MAP-RA effect was to improve precipitation fields
  for both models BOLAM and MOLOCH.
• In general explicit convection in MOLOCH allows
  a better representation of precipitation fields
  compared to the parametrized convection modelled
  at coarser resolution.
• More extensive tests needed in order to achieve
  more robust conclusions.

IOP-15
IOP-15

• Selected references
• Buzzi, A., DIsidoro, M. S. Davolio. A case
  study of an orographic cyclone formation south of
  the Alps during the MAP-SOP. Quart. J. Roy.
  Meteor. Soc, 2003a, 129, 1795-1818.
• Buzzi, A., Davolio, S., DIsidoro, M. P.
  Malguzzi. The impact of resolution and of 4-D VAR
  reanalysis on the simulations of heavy
  precipitation in MAP cases. Met. Zeitschrift,
  2003b, submitted.
• Keil, C C. Cardinali. The ECMWF Re-Analysis of
  the Mesoscale Alpine Programme Special Observing
  Period. ECMWF Internal Memor., 2003, 401, 34 p.

Contact Massimo DIsidoro ltm.disidoro_at_isac.cnr.i
tgt