Hi Thank you for your replies. Here is a summary (several have asked).
One useful link: http://www.ecmwf.int/newsevents/meetings/workshops/2011/MOS13/presentations/Hewson.pdf Bias is estimated, plume graphs and forecast models developed. One slide shows a statistically surprisingly narrow band (in red) normal plot for 42 forecast members. Lots of detail: variability for estimating signal greater than sampling error. 14 years is apparently enough to estimate linear bias correction. Hindcast and forecast determinations. See also: http://www.nws.noaa.gov/ost/climate/STIP/36CDPW/36cdpw-hkang.pdf by Kang et al. (2011). http://eurobrisa.cptec.inpe.br/3rdworkshop/alberto-arribas-barcelona.ppt (slide presentation) Ling Ling Huang Sacramento City College --- On Fri, 9/14/12, ling huang <[email protected]> wrote: From: ling huang <[email protected]> Subject: [ECOLOG-L] Predicting winter weather patterns (United Kingdom) To: [email protected] Date: Friday, September 14, 2012, 8:49 PM Hi I just finished reading a BBC article about a model predicting winter weather for the UK. The model is called GloSea4. The link is: http://www.bbc.co.uk/news/science-environment-19584302 I was interested in how they determine the model's confidence in predicting such events. If anybody knows of the model and the statistical procedures involved (I suppose some type of cross validation / jackknife etc.) please let me know. If others are interested I will write up a summary for the board. Best Ling Ling Huang Sacramento City College
