Sundog Posted January 7, 2011 Share Posted January 7, 2011 Elaborate on this for me if you could be so kind? If there were no observations assimilated at 18z, it would look effectively the same as 12z because of how data assimilation works (lagged six hours, so 120h from 18z would look like 126h from 12z). BTW, we STILL assimilate millions of observations during these cycles. I am curious, if a model cycle begins at say 6z, shouldn't all the subsequent runs improve verification wise if at least some fresh data is assimilated at all runs? Why aren't the verification scores in the same order as the runs, meaning 12z verifies better than 6z, 18z verifies better than 12z etc.? Link to comment Share on other sites More sharing options...
ezweather Posted January 7, 2011 Share Posted January 7, 2011 Just my opinion here. Sorry if I came across strong on this.. Probably better if I had said I use them for trends.. That would be it.. Anyway, back to tracking the storm for tomorrow and next week.. <BR><BR> Elaborate on this for me if you could be so kind? If there were no observations assimilated at 18z, it would look effectively the same as 12z because of how data assimilation works (lagged six hours, so 120h from 18z would look like 126h from 12z). BTW, we STILL assimilate <I><B>millions </B></I>of observations<I><B> </B></I>during these cycles. Link to comment Share on other sites More sharing options...
Amped Posted January 7, 2011 Share Posted January 7, 2011 That 18Z GFS was an odd run. It has a better looking ridge axis ahead of the western trough yet the shortwave sails through like the ridge isn't there and even weakens things a bit. Something isn't making sense. Yeah the troff won't dig into the southeast. You can see strong Negative vorticity zip into the southeast ahead of the storm on the PSU ewall sight. Link to comment Share on other sites More sharing options...
A-L-E-X Posted January 7, 2011 Share Posted January 7, 2011 Elaborate on this for me if you could be so kind? If there were no observations assimilated at 18z, it would look effectively the same as 12z because of how data assimilation works (lagged six hours, so 120h from 18z would look like 126h from 12z). BTW, we STILL assimilate millions of observations during these cycles. I wonder if anyone could answer this question I saw posed in the subforums-- why is it that the ensembles of a majority of the models are well to the NW of the OP runs? It isnt just the GFS, but the Euro and the GGEM that this is happening on. Looks like the s/w will be onshore by 0z so perhaps tonight or later tomorrow, things will fall more in line. Link to comment Share on other sites More sharing options...
dtk Posted January 7, 2011 Share Posted January 7, 2011 I am curious, if a model cycle begins at say 6z, shouldn't all the subsequent runs improve verification wise if at least some fresh data is assimilated at all runs? Why aren't the verification scores in the same order as the runs, meaning 12z verifies better than 6z, 18z verifies better than 12z etc.? You have to also consider lead times...what's important is that in a time-mean sense, a 120 hour forecast from 18z will verify better than a 126hr forecast from 12z. Generally speaking, forecast verification (given the same forecast lead time) is not different in a statistically significant sense between the four run times. Link to comment Share on other sites More sharing options...
A-L-E-X Posted January 7, 2011 Share Posted January 7, 2011 18z GFS ENSEMBLE MEAN is NW of the OP-- not really a surprise there. Link to comment Share on other sites More sharing options...
Paweatherguy1 Posted January 7, 2011 Share Posted January 7, 2011 18z DGEX location is even further west. Blows up once north of PHL Is it possible to post earlier frames?? Link to comment Share on other sites More sharing options...
baroclinic_instability Posted January 7, 2011 Share Posted January 7, 2011 Is it possible to post earlier frames?? It isn't because the DGEX is a NAM run extended by initializing it with GFS data. The DGEX essentially starts at 84 hours. Here is a little info. http://www.emc.ncep.noaa.gov/mmb/mmbpll/dgexhome.ops/ Link to comment Share on other sites More sharing options...
baroclinic_instability Posted January 7, 2011 Share Posted January 7, 2011 It isn't because the DGEX is a NAM run extended by initializing it with GFS data. The DGEX essentially starts at 84 hours. Here is a little info. http://www.emc.ncep....l/dgexhome.ops/ Sorry I see his image is at 108. http://www.meteo.psu.edu/~gadomski/DGEXEAST_18z/dgexloop.html Link to comment Share on other sites More sharing options...
Paweatherguy1 Posted January 8, 2011 Share Posted January 8, 2011 Sorry I see his image is at 108. http://www.meteo.psu.edu/~gadomski/DGEXEAST_18z/dgexloop.html Thanks. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.