Medium Range Discussion Winter 2011-12

[dtk]

The performance of the CAN based on that chart was baffling many of us. We dug to the bottom of it in another thread the other day and realized it was because a majority of the time the CAN has such a large spread between its ensemble members (looks like a squashed spider) it has a higher verification score. When you cover as much space on the phase diagram as the CAN does with its ensembles you're bound to "get it right"

That doesn't have to necessarily be true though since the deterministic skill that is calculated (on that chart) is for the ensemble mean (there are other verification metrics for EPS systems that take probability, ensemble spread, etc. into account, but I haven't seen such metrics used for MJO forecasting [as of yet] ).

I think an over-dispersive ensemble like the Canadian will only have the appearance of "better skill" during certain regimes (such as when things were relatively stagnant, and the other models were showing the MJO amplifying too quickly).

[dtk]

I don't think any model's verification of MJO prediction is worth the time that this forum has put into it. The basic 200mb structure has and continues to suggest that the eastern wave is legitimate and will alter the N PAC.

I don't really care if it nose dives in phase 8-1 to the COD. It doesn't mean anything because the damage has been done at that point. It isn't until the forcing reemerges in phases 4-5-6 again when we go back to winter 11-12 "default"...

Well said. People seem to put too much emphasis on the exact position/amplitude of the MJO and trying to connect that to the constructed analogs for sensible weather (there is a chart somewhere that shows the statistical significance of the constructed analogs for temperature and precipitation anomalies based on MJO phase. I'll post that if and when I find it).

[hm8]

(there is a chart somewhere that shows the statistical significance of the constructed analogs for temperature and precipitation anomalies based on MJO phase. I'll post that if and when I find it).

The right half of this chart, perhaps?

[TNE]

I'm guessing from the tone of this discussion that a cold snap no longer seems likely in February? JB and Joe D'Aleo spoke too soon?

Some things that you need to start understanding before the winter is over, 1. cold air has not been in this winter. 2. based on climo we are already past the coldest time of the year 3. there is opportunity for snow 4. about every fifteen days we get to observe cut-off or tail-end energy impacting the coast, this is important because it may trend toward the October event or toward a high precipitation event. 5. stop wishing for cold air. THERE IS 6. On the 12th-13th of February there is a chance for a winter storm.

We have a departing storm, which is part of a cold front. Not important. Then we have some leftover energy in the South (like we have seen throughout this whole freaking winter). Problem is it can trend in any direction. The energy does not phase, energy does phase and torch, energy phases and turns into a winter storm (hell yea.), energy does not phase and torches and we get a widespread rain event (this based on the GFS is unlikely). At this point the best two solutions are: it can phase and allow surface development (winter storm) or it can not phase and cause NOTHING. The surface is cold and is behind the cold front in the East (similar to how the October 29th event developed).

So here is the chance for the winter storm. The energies are close to phasing, if the phase occurs in this set-up, then we get a winter storm. Probably one of the best potentials I have seen this winter.

I did not monitor any other runs or models so I am not sure how this solution is trending. Now this is something to talk about, talking about cold air in this winter is borderline to wishcasting.

[DaculaWeather]

Well said. People seem to put too much emphasis on the exact position/amplitude of the MJO and trying to connect that to the constructed analogs for sensible weather (there is a chart somewhere that shows the statistical significance of the constructed analogs for temperature and precipitation anomalies based on MJO phase. I'll post that if and when I find it).

Or maybe this: http://www.daculaweather.com/seasonal_mean_olr_anomalies.php

[TNE]

Another nice aspect about Feb 12-13 and October 29 is the whole mjo thing. We had the mjo in a similar position and it was pretty active leading up to the event. Not sure if we had a Tropical Storm develop near Central America around Oct. 20 or not but this is good nonetheless. I am not sure we had this similarity with any other cut-off so far this winter.

p.s. Never over analyze something. post your thoughts once and leave it. Overanalyzing events usually causes error and is a complete waste of time if it does not work out.

[Ottawa Blizzard]

Some things that you need to start understanding before the winter is over, 1. cold air has not been in this winter. 2. based on climo we are already past the coldest time of the year 3. there is opportunity for snow 4. about every fifteen days we get to observe cut-off or tail-end energy impacting the coast, this is important because it may trend toward the October event or toward a high precipitation event. 5. stop wishing for cold air. THERE IS 6. On the 12th-13th of February there is a chance for a winter storm.

We have a departing storm, which is part of a cold front. Not important. Then we have some leftover energy in the South (like we have seen throughout this whole freaking winter). Problem is it can trend in any direction. The energy does not phase, energy does phase and torch, energy phases and turns into a winter storm (hell yea.), energy does not phase and torches and we get a widespread rain event (this based on the GFS is unlikely). At this point the best two solutions are: it can phase and allow surface development (winter storm) or it can not phase and cause NOTHING. The surface is cold and is behind the cold front in the East (similar to how the October 29th event developed).

So here is the chance for the winter storm. The energies are close to phasing, if the phase occurs in this set-up, then we get a winter storm. Probably one of the best potentials I have seen this winter.

I did not monitor any other runs or models so I am not sure how this solution is trending. Now this is something to talk about, talking about cold air in this winter is borderline to wishcasting.

you're right. I think the coldest Toronto might get is a high in the low 20s, which would be just slightly below average at this time of year. As for last October's event, that won't mean much to me given I saw nothing out of it. Would you believe that eitehr yesterday or the day before JB was apparently hinting at 0F day in Chicago and a 0F night in NYC? What is he seeing that nobody else is?

[Ottawa Blizzard]

you're right. I think the coldest Toronto might get is a high in the low 20s, which would be just slightly below average at this time of year. As for last October's event, that won't mean much to me given I saw nothing out of it. Would you believe that eitehr yesterday or the day before JB was apparently hinting at 0F day in Chicago and a 0F night in NYC? What is he seeing that nobody else is?

lol I might have to eat those words. The 18z GFS is showing bitter cold for next Sunday night!

[baroclinic_instability]

The Euro has never been particularly distinguished beyond D6...I recall seeing a verification chart back in 2008-2009 for the last 12 month period and it was almost exactly the same as the GFS at day 8....I think more than ever we have everyone looking beyond day 6 so people start remembering what it showed at day 8 or day 9 now whereas in the past we never even really looked at those panels...or certainly never took them with an ounce of credibility.

So when we see it show a b ig cold outbreak or a snowstorm at 216 hours and it doesn't materialize....we get the impression that the model stinks, when it used to do this all the time in the past, but we just had less data publically available to see it.

I would tend to agree with this. The verification scores for the past 2-3 decades pretty much tell the story...models have improved significantly and continue to improve (albeit at a slower pace). Personally I think a big problem in the 6-10 day is similar to issues we have to deal with in the shorter ranges. Models have become skillful enough that some tend to forget (and this includes many mets) models can be wrong--especially with extremes (big storms or major pattern regime shifts). The increasing use of ensemble modeling probably doesn't help that attitude as means can make it seem particular pattern regimes/changes are more probable. The reality is, as impressive as data assimilation is, the initial conditions models ingest are still a tiny fraction of the atmospheric volume, and they have their own errors associated with them.

[donsutherland1]

Another nice aspect about Feb 12-13 and October 29 is the whole mjo thing. We had the mjo in a similar position and it was pretty active leading up to the event.

FWIW, the MJO was in Phase 2 leading up to and during the October 29 snowstorm.

[donsutherland1]

lol I might have to eat those words. The 18z GFS is showing bitter cold for next Sunday night!

A fairly sharp but short-duration cold shot during the weekend (part of the February 8-15 timeframe) would fit the composite pattern based on teleconnections. I posted the chart earlier in this thread at: http://www.americanwx.com/bb/index.php/topic/29440-medium-range-discussion-winter-2011-12/page__st__771

I still do not believe the cold will lock in.

[GaWx]

Well said. People seem to put too much emphasis on the exact position/amplitude of the MJO and trying to connect that to the constructed analogs for sensible weather (there is a chart somewhere that shows the statistical significance of the constructed analogs for temperature and precipitation anomalies based on MJO phase. I'll post that if and when I find it).

I agree with you about this. I counted 17 episodes of MJO phases 8, 1 and 2 during Jan-early March since 1975. At Atlanta, six of the 17 were cold, eight of the 17 were near normal, and three were warm. So, whereas there would seem to be a slight cold tendency, the 17 averaged only slightly below normal average and more than half were near normal to warm.

More specifically, here are the 17 MJO phase 8, 1, and 2 periods analyzed for Atlanta with temp.'s and S/IP (above avg. S/IP for each period bolded):

-1/19-2/8/75: warm, T

-1/27-2/3/76*: norm, T

-2/7-28/78*: cold, 0.3"

-(1/20-30/83): cold, 1.9"

-1/12-2/3/85: cold, 0.4"

-2/21-3/16/88*: norm, 0.0"

-(2/7-23/90): warm, 0.0"

-1/13-2/2/92: norm, 5.0"

-1/3-15/93*: warm, 0.0"

-1/2-14/97: norm, T

-(1/6-12/98): norm, 0.0"

-2/16-24/99: norm, 0.4"

-1/10-27/04*: norm, T or 0.0"

-2/3-12/06: cold, T

-1/16-24/09: cold, T

-2/7-15/10: cold, 3.6"

-1/24-28/11: norm, 0.0"

* = strong MJO

() = rather weak MJO but still outside circle

Note that the three heaviest KATL S/IP's by far during these 17 MJO periods going through phases 8, 1, and 2 were all during El Nino's fwiw: 1/83's 1.9", 1/92's 5.0", and 2/10's 3.6". Also, four of the six El Nino's had measurable S/IP. In stark contrast, the eleven non-Nino's never had more than 0.4" and nine of the eleven non-Nino's had no measurable S/IP. So, that tells me that ENSO phase might very well be more of a factor than MJO phase as regards sig. S/IP chances.

It would be interesting to see how the analysis of these 17 periods would look for other major cities.

[Ottawa Blizzard]

A fairly sharp but short-duration cold shot during the weekend (part of the February 8-15 timeframe) would fit the composite pattern based on teleconnections. I posted the chart earlier in this thread at: http://www.americanw...2/page__st__771

I still do not believe the cold will lock in.

Don, from what I can make out from JB's twitter feed, he's saying that the stratospheric warming event is what led to the current cold wave in Europe and that he needs it (the cold wave) to happen in the US too, in order to prove his point that these warming events lead to cold periods a few weeks after they start. He seems to be implying that - in his view - we could be in for the February equivilent of what happened in January 1985 (Reagan inaugural outbreak), only 3-4 weeks later (making it mid February). We'll see, bearing in mind that the normals/averages are up from those of mid January. He's been calling for a flip since November though and has been busting big time.

[wxmx]

Looking at how poor it's handle the MJO recently, the ECMWF operational/ensemble 500mb forecasts beyond day 5 should be questionable. It's been trying to stall the MJO in phase 6, since Jan 31st.

This forecast from Jan 31st. It had:

Today, the MJO pulse is now strong entering phase 7. But ECMWF is still forecasting to collapse back into phase almost immediately again in phase 6:

I think I know whats going on with the Euro MJO forecasts...if you look at the first graph, it's the exactly same forecast as the second one...it hasn't updated!

[TNE]

February 5-13 vs October 21-29.

MJO:

Feb- STRONG. More progressive Phase 7-8.

Oct- STRONG. Phase 2.

What I do see though is a similarity in tropical development. We had Rita get noticed on October 21st, and for this period we have our very first Invest. Both impacting the East Coast as a weakened disturbance.

Upper Air Maps depict the pattern of the cut-off /tail-end energies throughout this winter:

February 11-13 2012-Snowstorm Potential

Cut-off (not as strong as we had it this winter) entering TX Panhandle at hr 123.

Cut-off Missing Phase..DAMN! This is why I do not like to follow storms until they hit. Try not to get excited folks.

NO PHASE. NO EVENT.

October 27-29 2011- Snowstorm

Cut-off (not as strong as we had it this winter) entering TX Panhandle on the 27th of October

PHASE. Historic Winter Storm.

Try to figure out how the tropics fit into this one. Invest 90L (East Coast) vs Rita (East Coast).

[The_Global_Warmer]

Heat Rises causing air to destabilize.

I was hesitant to post this stuff in this thread because it comes with the Global Warming headlines. But I must seek knowledge of this anomaly. It is the first anomaly of it's kind that we know of at the surface for the last 50 years. I don't care about why it's happening( that means no global warming discussion, but I care if there is a real causality, real correlation, or just random coinciding events. In-case you do not know, currently the Kara, Barents, Smidgen of the Arctic Basin, and Greenland sea are seeing there highest Sea Ice Area anomalies we have recorded. Inversely the Bering Sea is seeing the highest sea ice anomalies since the 1970s. However one out-weighs the other, so much so a new record low in sea ice coverage the last few days has been running, but this started back before this. All told to paint a more clear picture.

The combined extent anomaly between the four sea is currently running about 1.75-2.0 mil km2 below the 1979-2008 mean.

Here is where I get lost:

250MB.

400MB, the warmest anomalies are right where the ice becomes to thick for any heat release, but also where the WAA coming from the south would be mitigated. You see the anomalies spike where the absence of ice is on the climo record.

600MB, the signal is stronger and over the exact same spot as the 400mb signal.

850mb, obviously the same spot. And stronger.

925MB

Surface:

Anyways, there is quite a bit of heat coming out of the Ocean up there.

If there is no physical merit to this then please let me know. I have looked around and read some papers showing some plausibility to this. But I can't find anything that specifically throws a lot into this.

thanks

[thenewbigmack]

I normally post over at Westernwx, but I figured the composites I threw together would be more interesting for you guys anyways. These are similar to what Don Sutherland usually posts, and that is because him and I have been in contact for a while now. Just want to say thanks Don for all the help, very much appreciated.

Anyways, here we go. I took a look at the MJO progression, and aside from the Euro, all the models are in pretty good agreement. For those that don't have a link, here's the MJO chart I find the most useful:

http://www.daculaweather.com/mjo_phase_forecast.php

The below composites are made up from the above image, keeping region, amplitude, and month into consideration. The last image on the bottom-right is the GFS Ensemble analogs rolled forward 3 days, which matches up time wise with the last MJO composite

Pretty good agreement between the MJO composites and the GFS analogs that you guy's might get a chance for some winter weather. One thing that caught my eye today was the fact that the AO forecast had greatly increased compared to previous ones. If it were to verify, the chances and magnitude of cold may not be much at all, but we'll have to wait and see what future run's show. Take it easy everyone

[DaculaWeather]

Heat Rises causing air to destabilize.

I was hesitant to post this stuff in this thread because it comes with the Global Warming headlines. But I must seek knowledge of this anomaly. It is the first anomaly of it's kind that we know of at the surface for the last 50 years. I don't care about why it's happening( that means no global warming discussion, but I care if there is a real causality, real correlation, or just random coinciding events. In-case you do not know, currently the Kara, Barents, Smidgen of the Arctic Basin, and Greenland sea are seeing there highest Sea Ice Area anomalies we have recorded. Inversely the Bering Sea is seeing the highest sea ice anomalies since the 1970s. However one out-weighs the other, so much so a new record low in sea ice coverage the last few days has been running, but this started back before this. All told to paint a more clear picture.

The combined extent anomaly between the four sea is currently running about 1.75-2.0 mil km2 below the 1979-2008 mean.

Here is where I get lost:

250MB.

400MB, the warmest anomalies are right where the ice becomes to thick for any heat release, but also where the WAA coming from the south would be mitigated. You see the anomalies spike where the absence of ice is on the climo record.

600MB, the signal is stronger and over the exact same spot as the 400mb signal.

850mb, obviously the same spot. And stronger.

925MB

Surface:

Anyways, there is quite a bit of heat coming out of the Ocean up there.

If there is no physical merit to this then please let me know. I have looked around and read some papers showing some plausibility to this. But I can't find anything that specifically throws a lot into this.

thanks

Undersea fissures releasing heat? Plenty of volcanic activity there. The Eurasian and North American plates join in that region... actually they are diverging in that region. It could be a splitting of the plates is releasing magma undersea. Just guessing.

[The_Global_Warmer]

Undersea fissures releasing heat? Plenty of volcanic activity there. The Eurasian and North American plates join in that region... actually they are diverging in that region. It could be a splitting of the plates is releasing magma undersea. Just guessing.

Would that show up as OLR?

I never thought of that.

Most of it is likely from the gulf stream. And in situ heat from 65-40N north with the HPs bringing alot of sunny days.

more specifically from Feb 1st to the 5th the Kara sea had 40-60 w/m2+

This was when the ice was moved. the lid was taken off so to speak.

So my question is how will that affect the pattern? If at all

[Weathergun]

I think I know whats going on with the Euro MJO forecasts...if you look at the first graph, it's the exactly same forecast as the second one...it hasn't updated!

I thought this too. But the dates on second chart say otherwise. Has anyone seen the ECMWF MJO forecasts off the CPC site?

Edit: Never mind it's been corrected or updated.

[WidreMann]

I thought this too. But the dates on second chart say otherwise. Has anyone seen the ECMWF MJO forecasts off the CPC site?

It's very obviously the same forecast (look at the dots and lines -- they are a perfect match for the previous version), but on an updated map that has the progression since the old forecast. In other words, they messed up.

[phil882]

It looks like the ECMWF has finally been updated (or corrected) on the RMM plot. Seems like decently propagating MJO into phase 1/2...

[HM]

This MJO wave propagating more like the GFS ensembles became apparent a day or so ago when the IO convection dissapated and the h2 structure was clearly over the eastern most wave. This is legit folks!

What do you get when you get a MJO wave through phase 8-1-2 and a -AO in a La Nina February? I am sure Don has the composites already.

[wxmx]

This MJO wave propagating more like the GFS ensembles became apparent a day or so ago when the IO convection dissapated and the h2 structure was clearly over the eastern most wave. This is legit folks!

What do you get when you get a MJO wave through phase 8-1-2 and a -AO in a La Nina February? I am sure Don has the composites already.

Haven't seen it, but I'm plenty sure the composite map is full of blue for the eastern half of the CONUS.

[Chris L]

It is likely a good sign for those who want cold and snow. I just hope we can have a nice storm for all of us.

[am19psu]

Haven't seen it, but I'm plenty sure the composite map is full of blue for the eastern half of the CONUS.

Allen doesn't have it broken out by AO phase, but the Nina 1-2 composites are actually warm for the East. 8 is cold, though.

[wxmx]

Allen doesn't have it broken out by AO phase, but the Nina 1-2 composites are actually warm for the East. 8 is cold, though.

You have to look at things with a perspective. Is there any phase that's cold for the east other than phase 8 in a La Niña?

[am19psu]

You have to look at things with a perspective. Is there any phase that's cold for the east other than phase 8 in a La Niña?

Nope.

[HurricaneJosh]

Love it when (presumed) tropical dudes get into icekitty dirty dancing in public.

[am19psu]

Love it when (presumed) tropical dudes get into icekitty dirty dancing in public.

Don't you have some cold temperature obs from Prague to post somewhere?