Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

 

Section
Column

Plots from the Metview exercises

The gallery below shows example plots from all the practical exercises the participants were able to do with the Metview macros developed by Sandor Kertesz in ECMWF's forecast department.

Most of the plot types can be animated in time with Metview (not shown).

For more information, please contact: openifs-support@ecmwf.int.

 

 

Column
width30%
Panel
titleBGColorlightblue
titleOn this page...

Table of Contents
maxLevel1

Panel
borderColorblue
borderStyledotted

Analyses

(click image for full-size)

Image Added
Image Added
Image Added
Image Added
MSLP, Z500 and track of Hurricane NadinePV/320K and MSLPMSLP, 10m wind, Z500 & windspeed, SSTCross-section through Hurricane Nadine and cut-off low showing PV, theta and wind vectors. Illustrates warm core of Nadine and cold core of cut-off.

 

 

Panel
borderColorblue
borderStyledotted

HRES (deterministic) forecast

 

Image Added
Image Added
Image Added
MSLP & T2m, Z200 & windspeed,
equiv. potential temp., 850hPa wind & T
Total precipitationZ500 difference between HRES & analysis
Panel
borderColorblue
borderStyledotted

Ensembles

Image Added
Image Added
Image Added
RMSE curves Z500 for ensembleMSLP ensemble mean and spread
compared to analysis
MSLP ensemble spaghetti plot
Image Added
Image Added
Image Added
Ensemble stamp map total precipitation over FranceEnsemble difference stamp map of z500Comparing MSLP of perturbed ensemble member to analysis
Image Added
Image Added
 
Difference maps of two ensemble members to analysisCumulative distribution function MSLP for ensemble forecast
at Toulouse, France
 
Panel
borderColorblue
borderStyledotted

Principal component analysis and Clusters

After computing the EOFs a clustering can be applied to all of the above ensemble plots.

Image Added
Image Added
Image Added

EOF and PCA analyses of ensemble
spread for Z500 using clustering (top graph;
clusters shown as either red or blue dots).

Z500, ensemble mean of cluster 1 and 2
(top left & right respectively), compared
to HRES & analysis bottom.
Stamp difference map of MSLP with
clusters enabled. Maps are ordered
according to which cluster they are in
Image Added
Image Added
Image Added
Ensemble spaghetti plot with clusters. Top left and top right
show ensemble members in cluster 1 & 2 respectively.
Z500 RMSE curves with clusters.
Ensemble members in each cluster
are plotted separately.

Stamp map of total precip. with
clustering enable. Maps are ordered
according to their cluster.

Panel

Probabilities

Two ensemble forecasts datasets were provided for the workshop. The 2012 operational ensemble and a reforecast using the operational ensemble of 2016.

Image Added
Probabilities of total precipitation greater than 20mm
for both ensembles at 3 different forecast times.

Excerpt Include
Credits
Credits
nopaneltrue

Attendees

Students from ENM only - no other OpenIFS users this time.

Expect about 20-25 participants.

Accommodation

Panel
bgColorwhite
titleBGColorwhite
titleemail from Frederic
Concerning the accommodation it depends whether you want to be in Toulouse center or not.
If you want to be in Toulouse center there are plenty of choices : https://www.tripadvisor.fr/Hotels-g187175-Toulouse_Haute_Garonne_Midi_Pyrenees-Hotels.html
Maybe the Ibis and Mercure are cheaper than The Pullman... You'll have to take the metro and the bus to come to Meteo-France (approx 30 minutes)
If you want to be 2 minutes walk from Meteo France there is a B&B just next side of the road. But Meteo France neighborhood is not really amazing for a first visit in Toulouse...
I live in a small town called Cugnaux, at about 20 minutes from Toulouse. If you are interested I can catch you in the morning. Here is a nice address of a bed and breakfast, maybe a little expensive but with a huge breakfast in the morning : http://www.la-cognee.com/?lang=2

Topic

Storm Nadine. Case of extratropic transition & severe weather over France.

Panel
bgColorwhite
titleBGColorwhite
titleEmail from Frederic

Nadine seems to be a nice and recent case to illustrate predictability issues with OpenIFS. It would be great to use the SCM to illustrate the parametrization of convection. Half a day dedicated to Metview would also be useful. A colleague of mine could talk about Hymex field campain and CRM ensemble prediction of Mediterranean events with AROME. Maybe something about the clustering technique used by Pantillon too ?