Rainshadow Posted May 11, 2016 Report Share Posted May 11, 2016 BTW in our neck of the woods, no correction of the 2m temperature cold bias, especially over snow, an obvious .01" wet bias and lower stratospheric skill scores overall. 101 pages of happy reading: http://www.emc.ncep.noaa.gov/impdoc/GFS/CCBglobal2016.pdf 1 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted May 11, 2016 Author Report Share Posted May 11, 2016 Previous GFS day 5 skill scores was in 5th place trailing the JMA. The new version of the GFS(X), we are in 5th place, trailing the JMA. April 21-27 was very kind to modeling skills. After that, not so much... Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted June 4, 2016 Author Report Share Posted June 4, 2016 With qpf output like this: Its no wonder the current GFS has the worst verification skill scores as a decider between measurable and non-measurable pcpn of all these global models. 2 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted June 27, 2016 Author Report Share Posted June 27, 2016 Congrats on the 45 out of 48 hours of consecutive gentle rain the GFS gives PHL starting this evening, the 26th. 1 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted July 14, 2016 Author Report Share Posted July 14, 2016 This is a forecast for today (7/14), not day 13. Quote Link to comment Share on other sites More sharing options...
irishbri74 Posted July 14, 2016 Report Share Posted July 14, 2016 This is a forecast for today (7/14), not day 13. Untitled.jpg That price is right horn is pretty much the GFS theme song lol. Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted July 28, 2016 Author Report Share Posted July 28, 2016 I really hope there is improvement somewhere else, we have gone from let's forecast a 100 degree day somewhere to let's forecast nonstop rain. The 12z run on 7/28 has 27 consecutive and a second 42 consecutive hours of measurable rain at PHL in its first four days. Facepalm. Quote Link to comment Share on other sites More sharing options...
tombo82685 Posted July 28, 2016 Report Share Posted July 28, 2016 I really hope there is improvement somewhere else, we have gone from let's forecast a 100 degree day somewhere to let's forecast nonstop rain. The 12z run on 7/28 has 27 consecutive and a second 42 consecutive hours of measurable rain at PHL in its first four days. Facepalm. looking forward to it in winter when it forecasts 72 hrs straight of snow 1 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted July 29, 2016 Author Report Share Posted July 29, 2016 looking forward to it in winter when it forecasts 72 hrs straight of snow Here's a preview from 18z/28th GFS: it will not stop raining from Saturday morning through Monday evening. 60 straight hours of funfilled predicted wetness. Enjoy....... Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted July 29, 2016 Author Report Share Posted July 29, 2016 looking forward to it in winter when it forecasts 72 hrs straight of snow Hopefully you will see it snow out of the clear blue skies just like it will rain next week. This is 75 out of 84 hours or 25 out of 28 three hour segments with measurable rain starting Saturday. Quote Link to comment Share on other sites More sharing options...
rramblings Posted July 30, 2016 Report Share Posted July 30, 2016 I really hope there is improvement somewhere else, we have gone from let's forecast a 100 degree day somewhere to let's forecast nonstop rain. The 12z run on 7/28 has 27 consecutive and a second 42 consecutive hours of measurable rain at PHL in its first four days. Facepalm. But if we did get 60 hours of rain it could end the drought. Yeah, I know better be careful what I wish for . . . Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted July 30, 2016 Author Report Share Posted July 30, 2016 But if we did get 60 hours of rain it could end the drought. Yeah, I know better be careful what I wish for . . . You'll have to make it up on the backend ; the 27 consecutive hours of measurable rain in phase I was 9 hours of measurable rain spread over 16 hours. Quote Link to comment Share on other sites More sharing options...
dryslotted Posted July 30, 2016 Report Share Posted July 30, 2016 I really hope there is improvement somewhere else, we have gone from let's forecast a 100 degree day somewhere to let's forecast nonstop rain. The 12z run on 7/28 has 27 consecutive and a second 42 consecutive hours of measurable rain at PHL in its first four days. Facepalm. Hopefully this will help in the longer term (~3 year development window), though they've gone a different direction than the Euro from a core model perspective. Physics package and data assimilation scheme still tbd...It's an ambitious overhaul, and like many large, complex scientific computing/modeling projects, it has the potential to deliver with a thud, but I'll stay glass-half-full... Until then, we remain from run to run... https://www.washingtonpost.com/news/capital-weather-gang/wp/2016/07/27/the-weather-service-just-took-a-critical-first-step-in-creating-a-new-u-s-forecasting-model/ 2 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted July 30, 2016 Author Report Share Posted July 30, 2016 Hopefully this will help in the longer term (~3 year development window), though they've gone a different direction than the Euro from a core model perspective. Physics package and data assimilation scheme still tbd...It's an ambitious overhaul, and like many large, complex scientific computing/modeling projects, it has the potential to deliver with a thud, but I'll stay glass-half-full... Until then, we remain from run to run... https://www.washingtonpost.com/news/capital-weather-gang/wp/2016/07/27/the-weather-service-just-took-a-critical-first-step-in-creating-a-new-u-s-forecasting-model/ I was reading article, saddest part is only emergency fixes of current gfs as efforts in next gen gfs. I don't know enough about models to comment on Dr. Mass's comments. What Dr. Uccelini says is true, there is a point where the information gets out in time to be usable. The EC has better "timing" in Europe that a more prolonged processing procedure can be used. Also the EC as far as I know is not supporting/providing the background initial conditions for other models. 1 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted September 24, 2016 Author Report Share Posted September 24, 2016 Its a good thing the dew point (purple box) only reaches the mid 50s and 850mb dew points (red box) remain below zero C. Imagine how much additional pcpn would fall. As another aside, I dont know if I have ever seen 0C at 500mb observed or predicted before. Love those Tallahassee (this is 06z/Sept 24th run) dew points too. Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted March 23, 2017 Author Report Share Posted March 23, 2017 Gotta love it, 18z GFS has measurable pcpn occurring in 75 of 81 hours. Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted March 24, 2017 Author Report Share Posted March 24, 2017 Near 48 hours of non stop pcpn on the GFS from Saturday afternoon thru Monday afternoon, ducky.... Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted March 25, 2017 Author Report Share Posted March 25, 2017 19 of the next 20 3hr segments (12z March 25th) according to the GFS will be wet. Meanwhile, new nam not so much.... BTW next gen GFS (at least the 06z version) is not as continuously wet. It would be tougher to be wetter... Quote Link to comment Share on other sites More sharing options...
cbelke Posted March 26, 2017 Report Share Posted March 26, 2017 4 hours ago, Rainshadow said: 19 of the next 20 3hr segments (12z March 25th) according to the GFS will be wet. Meanwhile, new nam not so much.... BTW next gen GFS (at least the 06z version) is not as continuously wet. It would be tougher to be wetter... How long does it take to actually trust a model? Is the old model kept side by side with the new model? Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted March 26, 2017 Author Report Share Posted March 26, 2017 15 hours ago, cbelke said: How long does it take to actually trust a model? Is the old model kept side by side with the new model? I have read that there is general excitement about the new nam conus-wide. Although improvements are not the same everywhere. They (old vs new nam) were run in parallel for a while to compare and I believe one of the greatest improvements was with summer time convection, less of those flash flooding producing feedback bullseyes. As for the gfs and the new gfs, its really the next gen gfs that should show recognizable improvements I believe sometime later in 2018. For now we are stuck with pretty much what we have. The next implementation later this spring from what I read is only a slight improvement. 1 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted March 26, 2017 Author Report Share Posted March 26, 2017 Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted April 8, 2017 Author Report Share Posted April 8, 2017 The Sandy money cant come in to the GFS soon enough to get rid of these preposterous it is never going to stop raining runs it spews out every once in a while. (12z 4/7/17) Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted April 8, 2017 Author Report Share Posted April 8, 2017 Worst global model skill scores for light precipitation events across the conus last month.... When it comes to bias, there's the Euro and then there is all the rest. Quote Link to comment Share on other sites More sharing options...
Rainshadow Posted April 8, 2017 Author Report Share Posted April 8, 2017 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.