The March 25-March 32 HECS potential

[Ji]

looks like GFS fold to the euro. Euro wins again lol

[olafminesaw]

looks like GFS fold to the euro. Euro wins again lol

Hey now, the nam hasn't folded yet. We all know what that means, lol.

[olafminesaw]

Wow, the soundings aren't terrible 6z gfs even at hr.66. The timing also remains near perfect. If we could only get some heavier rates and a compliant EURO then we will be rocking.

[WinterWxLuvr]

Anybody gonna say they like the trends?

[Questsnow]

If the CWG isn't sticking to the original definition of the SPI, I think they should redefine it, then. It's really confusing people who actually know what it means.

 

The WPC has about an 80% chance of 2" or more for DC, by the way (though I think they rely too heavily on the GFS/NAM). Also, it probably doesn't take the 6z into account as it was issued at around 2 am.

http://www.hpc.ncep.noaa.gov/pwpf_48hr/prb_48hsnow_ge02_2013032212f072_sm.gif

 

[Quasievil]

This Storm is like groundhog day.

[WinterWxLuvr]

If the CWG isn't sticking to the original definition of the SPI, I think they should redefine it, then. It's really confusing people who actually know what it means.

The WPC has about an 80% chance of 2" or more for DC, by the way (though I think they rely too heavily on the GFS/NAM). Also, it probably doesn't take the 6z into account as it was issued at around 2 am.

http://www.hpc.ncep.noaa.gov/pwpf_48hr/prb_48hsnow_ge02_2013032212f072_sm.gif

I don't place any faith on those maps, but I do give credence to the forecaster maps. They were issued after the Euro and are pretty bullish.

[mitchnick]

Wunderground Euro snow maps show nothing for DCA/BWI

they were right on 3/6 when they showed nothing....just saying

[Questsnow]

Wunderground Euro snow maps show nothing for DCA/BWI

they were right on 3/6 when they showed nothing....just saying

They had nothing for Richmond, either. And areas down there got a few inches... just saying.

[mitchnick]

They had nothing for Richmond, either. And areas down there got a few inches... just saying.

I'll take the bet if you think they will be wrong on this event

you guys had the 5H low right over you, a meso feature that a global will rarely see

iow, you got lucky because the Wunderground Euro snow maps have been more right than wrong thru the season

[WxMan1]

If the CWG isn't sticking to the original definition of the SPI, I think they should redefine it, then. It's really confusing people who actually know what it means.

The WPC has about an 80% chance of 2" or more for DC, by the way (though I think they rely too heavily on the GFS/NAM). Also, it probably doesn't take the 6z into account as it was issued at around 2 am.

http://www.hpc.ncep.noaa.gov/pwpf_48hr/prb_48hsnow_ge02_2013032212f072_sm.gif

These probabilities are 3/4 SREF weighted, and are auto generated once the WWD forecaster sends his deterministic guidance out to the field internally (WFOs and RFCs).

http://www.wpc.ncep.noaa.gov/pwpf/about_pwpf_products.shtml

[usedtobe]

These probabilities are 3/4 SREF weighted, and are auto generated once the WWD forecaster sends his deterministic guidance out to the field internally (WFOs and RFCs).

http://www.wpc.ncep.noaa.gov/pwpf/about_pwpf_products.shtml

 

Brian,

 

My own perception is that more often then not, the product has the probabilities too high for the various thresholds. I'm not sure my perceptions are true or not.   Has Keith verified them so users can check to see how well they are  calibrated?  Wes

[WxMan1]

Hi Wes,

Yes, they have verified with a high bias over the years, but the perception is they will probably show better verification after this season given all the big events (particularly in the NE). I think overall they suffer more in the marginal events, given 1) the NAM's tendency (and thus the NMB members that make up 1/3 of the SREF) to be too wet into the cold air, and 2) when there are boundary layer concerns. Ironically, the WPC final 10/40/70% probabilities historically have shown little if any improvement over these prelim ones. Often times we see a 50% threshold and while that seems high, it would also indicate that 5/10 or 50/100 of such events would not verify statistically.

We have learned quite a bit over the past couple weeks via post analyses and collaboration with EMC in terms of the model's calculation of snowfall. It's actually kind of scary, honestly, during marginal BL events. I actually had an opportunity to present our (WPC's) perspective of the 3/6 event at EMC's weekly MEG meeting yesterday, highlighting these very issues. There are events as we've discovered on March 6th that there is a huge difference in marginal BL conditions between snow "fall" and the actual snow "depth". The models struggle with this...or as I should say (and concluded after the 3/6 storm)...WE as forecasters struggle sometimes with interpretating and differentiating between the various model snow accumulation output via various algorithms and techniques when the SFC and BL temps are so marginal. It's as if some of these maps represent snow that "would have" accumulated had the BL been sufficiently cold. The 'ol snow that melts once it falls dilemma.

We have seen much success with some of the experimental percent of frozen precip and rime-filtered SLR output from the NAM this past season. Again though, it's still the NAM, so if the NAM is too cold, then even these products will struggle as on 3/6. At least it can raise some red flags though when the BL is marginal, and/or when there is an elevated warm layer. Hopefully next year we can get this into the GFS, but right now it only works with Ferrier microphysics (i.e. NAM).

[Bob Chill]

Ohv transfers are bad. Bad for models. Bad for weenie nerves. And bad for dc snow. I'll stick with 1-3 for 95 West but now with daytime snow, elevation has gained a big upper hand. This doesn't include my yard.

Models are far from done. It can still go either way. History of events like this + calendar spring means the wrong way is favored.

[Ji]

00zeuro has been horrible the past 3 nights but rebounds at 12z

[WxMan1]

Ohv transfers are bad. Bad for models. Bad for weenie nerves. And bad for dc snow. I'll stick with 1-3 for 95 West but now with daytime snow, elevation has gained a big upper hand. This doesn't include my yard.

Models are far from done. It can still go either way. History of events like this + calendar spring means the wrong way is favored.

Monday is the start of spring break for my kids, and for a while now we've had a little mini vaca booked in the Smokies. If you can beat 'em, join 'em, although it will be an interesting trek along I-81 on Monday..

[dtk]

Hi Wes,

Yes, they have verified with a high bias over the years, but the perception is they will probably show better verification after this season given all the big events (particularly in the NE). I think overall they suffer more in the marginal events, given 1) the NAM's tendency (and thus the NMB members that make up 1/3 of the SREF) to be too wet into the cold air, and 2) when there are boundary layer concerns. Ironically, the WPC final 10/40/70% probabilities historically have shown little if any improvement over these prelim ones. Often times we see a 50% threshold and while that seems high, it would also indicate that 5/10 or 50/100 of such events would not verify statistically.

We have learned quite a bit over the past couple weeks via post analyses and collaboration with EMC in terms of the model's calculation of snowfall. It's actually kind of scary, honestly, during marginal BL events. I actually had an opportunity to present our (WPC's) perspective of the 3/6 event at EMC's weekly MEG meeting yesterday, highlighting these very issues. There are events as we've discovered on March 6th that there is a huge difference in marginal BL conditions between snow "fall" and the actual snow "depth". The models struggle with this...or as I should say (and concluded after the 3/6 storm)...WE as forecasters struggle sometimes with interpretating and differentiating between the various model snow accumulation output via various algorithms and techniques when the SFC and BL temps are so marginal. It's as if some of these maps represent snow that "would have" accumulated had the BL been sufficiently cold. The 'ol snow that melts once it falls dilemma.

We have seen much success with some of the experimental percent of frozen precip and rime-filtered SLR output from the NAM this past season. Again though, it's still the NAM, so if the NAM is too cold, then even these products will struggle as on 3/6. At least it can raise some red flags though when the BL is marginal, and/or when there is an elevated warm layer. Hopefully next year we can get this into the GFS, but right now it only works with Ferrier microphysics (i.e. NAM).

 

I enjoyed your presentation and I'm glad that the MEG has at least given us a forum for communication between EMC and some of the other centers (particularly you guys at WPC and SPC). 

 

It has become pretty obvious this winter that the "dominant ptype" stuff leaves a lot to be desired and we really need to look at other things.  Even in the 3/6 case where the NAM was too cold, there were clues that there were going to be issues (i.e. model output of % of frozen precip. being more in the 40-50% range).  Hopefully we can get the Ferrier microphysics working in the GFS, so we can start pursuing more sophisticated algorithms (especially as we push the global to higher resolutions).

[WinterWxLuvr]

Anybody have precip data for the Euro.  Seems that the issue (per Wund) with the Euro is the temps.  Seems to be quite a bit warmer than the other models.

[Ji]

Anybody have precip data for the Euro.  Seems that the issue (per Wund) with the Euro is the temps.  Seems to be quite a bit warmer than the other models.

 

its because there is little precipitation

 

jyo: .53 mostly during daytime

DCA. 50...temps 33-35.

 

its a disaster

[Ji]

the one good thing is the Euro Control which pretty much agrees with the Euro especially in 72 hour range is much more amped and gives us a good blow

[mattie g]

Can't wait for cold and drizzly.

 

What a great March.

[Deck Pic]

It's unfortunate that the best model is relegated to subscription status and even many subscriptions have crude output without a lot of panels. The NAM has really hurt forecasting in the winter going back since its inception. It does a bad job with pretty much every aspect of winter storms and to the extent forecasters blend it in even 5% will hurt forecasts. Though maybe it will be right this time. Afternoon temps at DCA on 3/25 of 29-30 degrees seems reasonable.

[Huffwx]

Great info, Wxman1. Thanks for taking the time to share.

[Ji]

It's unfortunate that the best model is relegated to subscription status and even many subscriptions have crude output without a lot of panels. The NAM has really hurt forecasting in the winter going back since its inception. It does a bad job with pretty much every aspect of winter storms and to the extent forecasters blend it in even 5% will hurt forecasts. Though maybe it will be right this time. Afternoon temps at DCA on 3/25 of 29-30 degrees seems reasonable.

well see if the GGEM upgrade has put it into elite status. This is a good test. It has been running slighty behind the ECMWF since its upgrade but blowing away the GFS. JMA is even beating the GFS

[mitchnick]

It's unfortunate that the best model is relegated to subscription status and even many subscriptions have crude output without a lot of panels. The NAM has really hurt forecasting in the winter going back since its inception. It does a bad job with pretty much every aspect of winter storms and to the extent forecasters blend it in even 5% will hurt forecasts. Though maybe it will be right this time. Afternoon temps at DCA on 3/25 of 29-30 degrees seems reasonable.

NAM performed well in 09/10

I'm guessing it has certain atmospheric conditions where it excels and fails

it certainly sux in a NINA so maybe it will prove itself as competent if we can ever get a NINO again

[Deck Pic]

well see if the GGEM upgrade has put it into elite status. This is a good test. It has been running slighty behind the ECMWF since its upgrade but blowing away the GFS. JMA is even beating the GFS

Verification scores don't mean a whole lot to me when it comes to winter storms in my backyard. I know how the models perform. I've been a much better forecaster pretending the NAM doesn't exist, which is a problem. And I am not a model basher. I am blown away by what models can do. I also understand it is a funding issue too. Apparently it is good with convection and squall lines so it does have a role. I just know in the winter it is wretched for winter storms in our area. There is a reason I was starting Nam- less threads several winters back. Because it sucks. And the Srefs suck too

[Yeoman]

Screw the NAM and invest in the CFS

[Deck Pic]

NAM performed well in 09/10

I'm guessing it has certain atmospheric conditions where it excels and fails

it certainly sux in a NINA so maybe it will prove itself as competent if we can ever get a NINO again

All the models are worse in Nina's for our backyard in winter storms. Most models were good in 2009-10. A wet cold over amped model is going to do pretty well in a 70-100" winter.

[WinterWxLuvr]

Verification scores don't mean a whole lot to me when it comes to winter storms in my backyard. I know how the models perform. I've been a much better forecaster pretending the NAM doesn't exist, which is a problem. And I am not a model basher. I am blown away by what models can do. I also understand it is a funding issue too. Apparently it is good with convection and squall lines so it does have a role. I just know in the winter it is wretched for winter storms in our area. There is a reason I was starting Nam- less threads several winters back. Because it sucks. And the Srefs suck too

 

I'm coming around to this conclusion myself.

 

Too bad the GFS is so fickle as we get close to winter events.  Instead of becoming more consistent with its forecasts (as we get closer), it seems to become more erratic.  The problem for many of us wrt using the Euro for winter ideas is that the access is so limited it's as if it doesn't even exist.  So, we are left with the GFS, NAM, and SREFS.  Probably be best just to wait until Sunday night and look out the window.

[Deck Pic]

well see if the GGEM upgrade has put it into elite status. This is a good test. It has been running slighty behind the ECMWF since its upgrade but blowing away the GFS. JMA is even beating the GFS

In the 11 years I have been following models I haven't used the GGEM or RGEM a single time ever so I can't really speak to it