Versions Compared
Key
- This line was added.
- This line was removed.
- Formatting was changed.
Introduction
A forecast 'bust' is typically associated with a sudden and pronounced drop in forecast performance. The standard way of assessing the performance of a single forecast is to calculate how well the height of the 500-hPa pressure field (Z500) agrees with the observed outcome. The agreement is typically quantified by calculating the area-averaged root-mean-square error (RMSE) or the spatial anomaly correlation coefficient (ACC).
This case study looks at an example of a 'forecast bust' from April 2011 over Europe. The reference for this case study is:
Panel | ||
---|---|---|
| ||
Rodwell, M.J. and Coauthors, (2013). Characteristics of occasional poor medium-range weather forecasts for europe. Bull. Am. Meteorol. Soc., 94(9) 1393–1405. DOI: http://dx.doi.org/10.1175/BAMS-D-12-00099.1 Rodwell, M. J., and Coauthors, 2012a: Characteristics of occasional poor medium-range forecasts for Europe. ECMWF Newsletter, No. 131, ECMWF, Reading, United Kingdom, 11–15. |
Case study aims
Possible approaches to studying this case are:
- Characteristics of forecast bust compared to analyses.
- Role of organised convection over the USA.
- Understanding Rossy wave development associated with the forecast bust.
- Use of ensembles.
- Sensitivity to parametrized physics processes, particularly vertical mixing.
Initial conditions
Case study initial conditions are provided on the OpenIFS ftp site. The ftp site is password protected, only licensed institutes may be provided with the ftp password. Please contact openifs-support@ecmwf.int.
The initial conditions are available for different resolutions and start dates. The experiment ids listed below are created at ECMWF and used for identifying the model forecasts on the ECMWF archive system (for those with access).
Note that ERA-Interim has a horizontal resolution of T255.
Info |
---|
As OpenIFS is a spectral model, the 'T' number refers to the triangular truncation in spectral space. Equivalent grid resolutions are: The number of vertical levels is given after the letter 'L' e.g. L62 means 62 vertical levels. Please note that higher resolutions progressively require more processors and computer memory to run. |
Panel | |||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||||||||
|
Acknowledgements
The figures on this page were prepared by Lauri Tuppi, University of Helsinki, who completed a project on this case study.
HTML |
---|
<script type="text/javascript" src="https://software.ecmwf.int/issues/s/en_UKet2vtj/787/12/1.2.5/_/download/batch/com.atlassian.jira.collector.plugin.jira-issue-collector-plugin:issuecollector/com.atlassian.jira.collector.plugin.jira-issue-collector-plugin:issuecollector.js?collectorId=5fd84ec6"></script>
|
Excerpt Include | ||||||
---|---|---|---|---|---|---|
|
Panel | ||||||
---|---|---|---|---|---|---|
| ||||||
Figure shows ACC (%) of a 6 day forecast initialized on dates shown. (courtesy Lauri Tuppi, University of Helsinki). |