WG4: Data Assimilation and
Predictability
George C. Craig
DLR-Institut für Physik der Atmosphäre
Institut für Physik der Atmosphäre
Goal: To quantify and extend the limits of predictability of convection
through high-resolution ensemble forecasting and advanced data
assimilation.
Scientific Questions:
DAP 1. What are the relative roles of upper and mid-tropospheric forcing versus
local orographic and surface flux influences on the predictability of convective
precipitation in a region of moderate orography?
DAP 2. What is the impact of the assimilation of high-resolution remote sensing data
on short-range forecasts of convective precipitation? Can assimilation help to
improve process understanding and model parameterizations?
DAP 3. What is the impact of model errors on forecast accuracy, in comparison to
error in initial fields, and can a synergetic use of observations lead to a
characterization and reduction of model error?
DAP 4. Is there an obvious impact of COPS measurements on model forecasts?
DAP 5. If reasons for lack of positive impact can be identified, can a better
measurement strategy be devised?
DAP 6. Which assimilation systems best handle the data and which may be practical
for real-time use (nudging, 3DVAR, 4DVAR, ensemble-based)?
DAP 7. Is such a system a valuable tool to support mission planning?
Institut für Physik der Atmosphäre
Folie 3 > Vortrag > Autor
Plans
Real-time
•
•
•
•
ETReC 2007
University of Hohenheim
DLR - COSMO-LEPS
Operational products from national meteorological services
Near real-time
•
Sensitivity studies
Post-Campaign Studies
Institut für Physik der Atmosphäre
Folie 4 > Vortrag > Autor
Episode I: 9 July 2002
Meteosat 7 IR 16:00 UTC
Lokal Modell: all 10 clusters
Institut für Physik der Atmosphäre
Folie 5 > Vortrag > Autor
Image matching as pre-conditioner for DA
1. COSMO-LEPS ensemble forecast highly non-Gaussian error stats.
2. Image matching selects
best sub-ensemble Gaussian error stats.
3. Conventional DA
assimilates most recent data
for short-range forecast
Institut für Physik der Atmosphäre
Folie 6 > Vortrag > Autor
Preconditioning for Latent Heat Nudging
(Daniel Leuenberger, MeteoSchweiz)
Institut für Physik der Atmosphäre
Folie 7 > Vortrag > Autor
DLR Realtime
•
•
•
Image matching FQM in real-time COSMO-LEPS
Provision to operations center
Near real-time impact studies
– impact of observations
– impact of preconditioning
Institut für Physik der Atmosphäre
Folie 8 > Vortrag > Autor
ETReC 2007
Goals
1. Predictability
2. Adaptive Observing System
Timeline
T+2-3 days: Calculation of sensitive regions for two target regions
– COPS
– MEDEX/HYMEX
T+1-3 days: Targetted observations (T+24)
– EURORISK-PREVIEW - radiosonde, AMDAR, ASAP
– COPS - airborne lidar
T+0.5-1 day: Improved forecasts
– NMS with GTS data
– COPS real-time/near real-time
Institut für Physik der Atmosphäre
Folie 9 > Vortrag > Autor
27 June 2006: 06 UTC – 12 UTC
Vortragstitel
10
Méso-NH Simulation :
Initial Time 27-06-06 06 UTC
3h accumulated precipitation (09-12 UTC) :
ICE 4 /
ARPEGE
ICE 3 / ECMWF
ICE 3 /
ARPEGE
Vortragstitel
11
Initial fields : Horizontal wind @ z =1
km
ARPEGE
ECMWF
Vortragstitel
12
Targeted observations in sensitive
areas
TTT–––12
24hhh
36
Verif. time: 27 june 12UTC
Obs. time: 26 00UTC; 12UTC, 27 00UTC
Vortragstitel
EPS: 25 June 12UTC (Courtesy of P. Arbogast)
13
Issues
•
•
•
•
Data exchange and archive
Data for real-time use
– GTS
– cutoff times
Provision of products for operations centre
Participation of national meterological services – ETReC 2007 meeting Exeter 11 Oct. 2006
Institut für Physik der Atmosphäre
Folie 14 > Vortrag > Autor
Initial fields: Relative humidity @
Z=1km
ECMWF
ARPEGE
Vortragstitel
15

Lorem ipsum dolor sit amet consectetuer Quis nostrud exerci tation