From: Tim Hewson <tim.hewson@ECMWF.INT>
 Subject: Re: Post-blizzard navel gazing Take 4: Optimal extraction of ensemble information
 Date: 28 January, 2015 at 18:15:36 GMT
 To: <MAP@LISTSERV.ALBANY.EDU>
 Reply-To: Tim Hewson <tim.hewson@ECMWF.INT>

 Gary and everyone,

 Nice to read all the posts on this event. This morning I even used the case as a discussion point in a training course lecture on 'forecast jumps'!

 Attached is an example plot that tries to summarise EC Ensemble handling of the event. It depicts 700 forecast and 500 re-forecast (M-Climate) runs in CDF format. On close examination a lot can be learnt. I'll just highlight one thing. The 24-48h forecast for NYC (darkest blue line) would imply that "all snow depths between 5 and 26ins are equally likely" *.

 Tim

 *using a 12:1 ratio