Second SRNWP Workshop on Statistical and Dynamical Adaptation

5-7 May 2003, Vienna, Austria

Participants

Stjepan Ivatek-Sahdan, Martina Tudor, Juha Kilpinen, Frederic Atger, Volker Renner, Philippe Crochet, Fabrizio Ciciulla, Fabrizio Nerozzi, John Bremnes, Malgorzata Melonek, Otilia Diaconu, Antonio Mestre, Anders Persson, Pierre Eckert, Danial Cattani, Jean Quiby, Wim de Rooy, Bodo Ahrens and Thomas Haiden.

find the agenda (with presentations) and the meeting report.

1. Agenda and presentations

Monday, 5 May 2003
09:00-09:20	Steinhauser, P., T. Haiden	Opening, organizational matters
09:20-09:30	J. C. Quiby,	A few words from the SRNWP Coordinator
Session 1: Statistical / probabilistic methods
09:30-09:55	Bremnes, J. B.	Probabilistic forecasts of precipitation in terms of quantiles ⬇
09:55-10:20	Persson, A.	Better forecasts yield larger errors? Experiences of statistical adaptation of deterministic and ensemble forecasts of 2 m temperature ⬇
10:20-10:45	Theis, S., A. Hense, U. Damrath, and V. Renner	Statistical postprocessing of weather parameters for a high-resolution limited-area model ⬇
10:45-11:10	Coffee Break
11:10-11:35	Melonek, M.	Windspeed prediction improvement by the MOS technique in windfarm case
11:35-12:00	Diaconu, O.	MOS based on the ALADIN numerical model ⬇
12:00-13:00	Lunch Break
13:00-13:10	Weather briefing
13:10-13:35	Kilpinen, J.	The development of statistical interpretation and adaption at Finnish ⬇
13:35-14:00	Farges, S., I. Souyri, M.-H. Théron, and F. Atger	Statistical adaptation for the prediction of unfrequent meteorological events ⬇
	Bergot, Th. and J. Noilhan	Fog Forecasting at Roissy Airport (Paris) with 1D model ⬇
	14:00-14:25	Forecasting extreme meteorological events over complex topography I:
	Eckert, P.	Pattern recognition by Artificial Neural Networks (ANN) ⬇
14:25-14:50	Forecasting extreme meteorological events over complex topography II:
14:25-14:50	Cattani, D.	Extreme Forecast Index (EFI) and Limited Area EPS (LEPS) ⬇
14:50-15:15	Coffee Break
15:15-15:40	Crochet, Ph.	Quantitative precipitation forecast mapping in Iceland using topographic information ⬇
15:40-16:05	Paccagnella, T., C. Marsigli, A. Montani, F. Nerozzi, and S. Tibaldi	QPF probabilistic prediction by a limited area model ensemble prediction system
19:00	Workshop Dinner
Tuesday, 6 May 2003
Session 2: Physical / dynamical & combined methods
09:00-09:25	Jackson, S. D., T. J. Wheeler, and W. P. Hopwood	The Met Office Road Surface Temperature Model (MORST)
09:25-09:50	Bergot, Th., D. Carrer, and J. Noilhan	France fog forecasting at Paris airport with 1D-model ⬇
09:50-10:05	Ciciulla, F.	A deterministic post-processing program applied to the Local-Model output fields, part 1 and 2
10:05-10:30	Coffee Break
10:30-10:55	Quiby, J. C.	Computation and operational production of trajectories from a meso-scale NWP model ⬇
10:55-11:20	Ivatek-Sahdan, S.	Operational use of the dynamical adaptation for high-resolution ALADIN forecast in the Dinaric Alps ⬇
11:20-11:45	Tudor, M	Use of dynamical adaptation in research impact studies ⬇
11:45-12:10	de Rooy, W., and K. Kok	A combined physical/statistical approach for the downscaling of wind speed from a NWP model ⬇
12:10-13:10	Lunch Break
13:10-13:20	Weather briefing
13:20-13:45	Ahrens, B., and A. Beck	Comparison of dynamical and stochastical downscaling in the framework of ALADIN ⬇
13:45-14:10	Haiden, T., S. Greilberger, and A. Schmalwieser	T2m nowcasting: statistical vs. physical adaptation ⬇
14:10-15:00	Final discussion

2. Meeting Report

Jean Quiby, SRNWP Network Coordinator

Is statistical adaptation still necessary?

Or as it has been said to the participants: What would you answer to your Director if he would ask you to work now to the improvement of the model and no longer for the adaptation?

Some answers:

Models are improving but are not yet perfect: it is thus perfectly justified to continue to work to the improvement of the model results.
We are now asked to make very local weather forecasts. Today's model resolutions cannot take into account very local effects. Thus we must relay on statistical adaptation.
If you want to correct a weakness in the results of a numerical model, you can do it much more rapidly with a statistical correction than by changing the model formulation.
Some needed parameters are not computed in the model. Example: visibility. This has to be done in the post-processing.

Assessment of the model results: are we using the right scores?

Is rmse a good score? Maybe, but one thing is sure: it must only be used with other scores. The reason is well known: it rewards smooth, undetailed results (as 2D-fields) and punish the scientist who does work at high resolution or develops schemes prone to handle extreme situations (cf. presentation of Anders Persson).

If rmse is used for the verification of 2D-fields, it must be accompanied by the variance. Thus very smooth fields, which will have low rmse, will be punished with the variance.

It has been said that for point verification (normally an observing station) we use to many scores. After definition of thresholds, "hit rate" and "false alarm rate" should be sufficient.

Probability forecasts

The works presented were based on the ECMWF ensembles of 51 members. Participants have been asked whether they have remarks about the ECMWF products. Two remarks have been made:

51 members with a good spread are enough. No need is seen for more members (the Centre plans to increase the ensemble to 100 members)
a colleague thinks that the deterministic high resolution 12 UTC integration should be suppressed and the effort put on an increase of the resolution of the ensemble.

Next meeting

A colleague proposed to hold this workshop annually instead of every two years. But the majority of the participants who aired their opinion on that point clearly favored a 2-year cycle. The next meeting will take place in May 2005 in Vienna.