Winter 2013-2014

[Rainshadow]

What I'm getting at is maybe delta T is more predictive than absolute T. But maybe I'm totally wrong.

Its possible.  October is warming at a slower rate than most months, so hoping that its feeling less of the effects. Also in the past beyond the last twenty or thirty years warm Octobers have shown similar results in lets say a colder overall environment. I have had problems with November because its warmed so much that I am looking at some :"normal months" if October is either normal or warm.  I'm sorry this cant be made smaller, but this is the October scoreboard the last 10 years.  This is the entire temperature database going back to 1872 divided into thirds.  First is October, second descriptor is ensuing winter.

 

[Rainshadow]

Snowlurker,

 

They fixed the web site, October has had the slowest rate of rise than any month (February is first); November has been faster.

 

 

[donsutherland1]

With the Philadelphia area likely to see its first snowflakes overnight into tomorrow morning with perhaps some coatings on grassy surfaces, I thought it would be a good idea to take a look at AO/PNA combinations for Philadelphia's 6" or greater snowstorms. The below charts are based on months where the ENSO Region 3.4 anomaly was greater than -0.5°C and less than +0.5°C. The charts reflect all such snowstorms that occurred during the above ENSO conditions from January 1950 through October 2013.

 

 

The AO-/PNA+ combination saw almost half of such snowstorms take place. The AO-/PNA+ saw the least such storms.

[snowlurker]

Snowlurker,

They fixed the web site, October has had the slowest rate of rise than any month (February is first); November has been faster.

Untitled.jpg

Untitled2.jpg

So much for my idea

[Rainshadow]

So much for my idea

Regardless, its going to be at least 11 years since the last "cold" October, it would be nice to see that side of the equation work.

[Rainshadow]

With the Philadelphia area likely to see its first snowflakes overnight into tomorrow morning with perhaps some coatings on grassy surfaces, I thought it would be a good idea to take a look at AO/PNA combinations for Philadelphia's 6" or greater snowstorms. The below charts are based on months where the ENSO Region 3.4 anomaly was greater than -0.5°C and less than +0.5°C. The charts reflect all such snowstorms that occurred during the above ENSO conditions from January 1950 through October 2013.

 

 

The AO-/PNA+ combination saw almost half of such snowstorms take place. The AO-/PNA+ saw the least such storms.

Don,

 

Good info as always!   While enso neutral positive winters are half as common as enso neutral negative winters, the historical difference in the number of large single events (>6") at PHL is huge: 2 vs 13.

[chubbs]

Bolaris calling for a big winter with  60/61 and 78/79 analogs

 

http://www.philly.com/philly/news/John_Bolaris_winter_weather_outlook.html

[HM]

Interesting, that winter is one of the stratospheric analogues that over in the UK that we would like to see repeated. And even though west QBO SSW's are notoriously difficult beasts to achieve as Jan 1991 and 2009 have shown. With the El Nino that isn't and rate of snow advance increase on the right side of average and the low sea ice years giving a positive high latitude H500 anomaly, I am thinking that a SSW may still be achievable this winter.

 

nhtime-1year.png

 

The 1991 SSW was a classic split - so we would need to see an increase in wave 2 activity later in winter for something similar, perhaps like 1979 also  ( another analogue comparable year)

 

1991 SSW

 

http://curriculum.pmartineau.webfactional.com/wp-content/svw_gallery/test/gif/1991_01_28.gif

 

Feb 1979 SSW

http://curriculum.pmartineau.webfactional.com/wp-content/svw_gallery/test/gif/1979_02_27.gif

 

 

I think that we should look out for irriversible deformation of the polar vortex in January where air with high PV is picked off the vortex and mixed into the mid latitudes - this could be a precurser of a later SSW. This process is described in this paper which includes the 1979 SSW PV charts.

 

http://www.columbia.edu/~lmp/paps/waugh+polvani-PlumbFestVolume-2010.pdf

 

I also think that the Arctic low sea ice years will contribute to the tropospheric feedback - with increased amplitude Rossby waves more likely to break into the strat. There is quite a considerable H500 anomaly difference between the high and low sea ice years which should be 'weighted' into any winter thoughts.

 

High sea ice anomaly

 

High sea ice Dec - Feb.png

 

 

Low sea ice anomaly

 

Low sea ice Dec-Feb.png

 

This year's diabatic circulation is still better than the coldest strat years with no SSW. I'm not ruling a SSW out, especially with the frequency increase we've seen over the last decade and the propensity for a N PAC low. Notice again the forcing will be shifting into the W-C Pacific and possibly bring another round of "el nino-like" flow late Nov-early Dec. Finally modeling and other vendors are catching on to this pattern shift.  

 

As for your sea ice anomaly charts, I do always appreciate mirror images. However, how did you adjust for the PDO which may be the signal you are getting there over the sea ice?

[donsutherland1]

Don,

 

Good info as always!   While enso neutral positive winters are half as common as enso neutral negative winters, the historical difference in the number of large single events (>6") at PHL is huge: 2 vs 13.

Thanks Tony.

 

Do you have any thoughts on why that might be the case? Perhaps it's localized (I'll have to check Washington's and New York's data tonight to see if the same holds true in one or both cities, even if the difference is not as extreme as in Philadelphia). Perhaps sample size plays a role.

[chubbs]

Don,

 

Good info as always!   While enso neutral positive winters are half as common as enso neutral negative winters, the historical difference in the number of large single events (>6") at PHL is huge: 2 vs 13.

Wonder if trend has any impact.  Looks like we could cross over from nuetral neg to nuetral pos sometime this winter.

[Rainshadow]

Thanks Tony.

 

Do you have any thoughts on why that might be the case? Perhaps it's localized (I'll have to check Washington's and New York's data tonight to see if the same holds true in one or both cities, even if the difference is not as extreme as in Philadelphia). Perhaps sample size plays a role.

 

Don,

 

It could be sample size; 6 years vs 13 years. Its not that they were single digit winters, just less big events here.  One of them was 93-94. so big snow difference for us vs NYC.

[iceman56]

Bolaris calling for a big winter with  60/61 and 78/79 analogs

 

http://www.philly.com/philly/news/John_Bolaris_winter_weather_outlook.html

"we are in an extreme weather cycle that will last through our lifetime and beyond. The globe is in a blow-torch state, with the warming of our oceans and atmosphere continuing to lead to unprecedented superstorms."

 

Hmmm....

[chubbs]

"we are in an extreme weather cycle that will last through our lifetime and beyond. The globe is in a blow-torch state, with the warming of our oceans and atmosphere continuing to lead to unprecedented superstorms."

 

Hmmm....

Sounds like he is basing analog selection mainly on ENSO state. Not sure other variables match.

[GD0815]

Bolaris calling for a big winter with  60/61 and 78/79 analogs

 

http://www.philly.com/philly/news/John_Bolaris_winter_weather_outlook.html

His forecast is predicated on a negative AO for the majority of winter, an assumption which I have not seen much support for.

[donsutherland1]

His forecast is predicated on a negative AO for the majority of winter, an assumption which I have not seen much support for.

From the article, it seems he favors an AO-/PNA+ in combination with the development of weak to possibly moderate El Niño conditions during the winter. It will be interesting to see how things play out, but those assumptions are key to his 30"-40" snowfall forecast.

 

Long-range forecasting is very difficult. Forecasting seasonal snowfall is even more difficult than temperature forecasts.

[atownwxwatcher]

Bolaris calling for a big winter with  60/61 and 78/79 analogs

 

http://www.philly.com/philly/news/John_Bolaris_winter_weather_outlook.html

Not sure when looking at the Enso plumes that I am seeing what he is seeing in reference to the El Nino. That seems to be the main focus of his heavier projection for snowfall totals besides the AO and PNA...

[GD0815]

From the article, it seems he favors an AO-/PNA+ in combination with the development of weak to possibly moderate El Niño conditions during the winter. It will be interesting to see how things play out, but those assumptions are key to his 30"-40" snowfall forecast.

 

Long-range forecasting is very difficult. Forecasting seasonal snowfall is even more difficult than temperature forecasts.

Don,

 

For sure...snowfall totals can be completely skewed by one storm--I really only look at them as educated guesses.  I think temperature forecasts have more skill in the long range, but I know even that is far from an exact science.

[chubbs]

He is going out on a limb. Out of last 129 winters  phl has had over 30" 25 times and between 30" and 40" only 14 times.

[donsutherland1]

Don,

 

For sure...snowfall totals can be completely skewed by one storm--I really only look at them as educated guesses.  I think temperature forecasts have more skill in the long range, but I know even that is far from an exact science.

I completely agree. From my observations, it seems that temperature forecasts can have some skill. Precipitation and snowfall ones have not yet reached that level. The level of complexity is greater for the precipitaton forecasts, as one is dealing with variables such as storm tracks, timing in relation to the cold air, etc. My guess is that even when one considers a degree of uncertainty/chaos, there is ample room for improvement in all those areas in the future from a combination of a better understanding of processes, improving technology on which models/ensembles are run, and empirical testing of forecasting approaches.

[donsutherland1]

He is going out on a limb. Out of last 129 winters  phl has had over 30" 25 times and between 30" and 40" only 14 times.

I agree. But his assumptions (El Niño + AO-) would provide a greater chance at such an outcome. Here is the data since 1950-51 when regular data on the AO was kept:

 

Total winters: 63

Winters with 30" or more snow: 12

19.0% had 30" or more snow

 

Winters with a DJF average AO of -1 or below: 17

Winters with 30" or more snow: 6

35.3% had 30" or more snow

 

Winters with a DJF average AO above -1: 46

Winters with 30" or more snow: 6

13.0% had 30" or more snow

 

Winters with a DJF average AO above 0: 23

Winters with 30" or more snow: 1

4.4% had 30" or more snow (1983 blizzard)

 

Winters with DJF average AO of +0.25 or above: 18

Winters with 30" or more snow: 0

0.0% had 30" or more snow:

 

Winters with a DJF average AO of -1 or below and ENSO R3.4 anomaly of +0.5°C or above: 7

Winters with 30" or more snow: 3

42.9% had 30" or more snow

 

Note: Sample sizes shrink, creating additional uncertainty.

 

In short, if one is able to reasonably estimate whether there will be strong, weak, or very little blocking, one has some idea about the probability of a very snowy winter (30" or above), but that's really about it. The majority of even strong blocking winters does not produce blockbuster 30" or above seasonal snowfall.

 

If I were trying to make an estimate, I'd refrain from providing a specific snowfall estimate or range. I might suggest an above climatalogical probability of high/low snowfall.

 

Unlike Mr. Bolaris, my early thinking favors an AO+ winter (SAI, OPI, and my own statistical probabilities). If so, prospects for 30" or greater snowfall would appear to be quite low. (I hope I'm wrong). 

[chionomaniac]

This year's diabatic circulation is still better than the coldest strat years with no SSW. I'm not ruling a SSW out, especially with the frequency increase we've seen over the last decade and the propensity for a N PAC low. Notice again the forcing will be shifting into the W-C Pacific and possibly bring another round of "el nino-like" flow late Nov-early Dec. Finally modeling and other vendors are catching on to this pattern shift.  

 

As for your sea ice anomaly charts, I do always appreciate mirror images. However, how did you adjust for the PDO which may be the signal you are getting there over the sea ice?

I didn't - which I guess is your point. But even though there is a predominance of +PDO in the high sea ice years and -PDO in the low sea ice years, there is still enough variability to show some sort of signal. Sadly there is a lack of years where the sea ice has reduced combined with a +PDO to formulate a true meaningful anomaly composite. 

 

However to edit, if we look at the mainly -PDO years that the low sea ice years encapsulate and compare them to the -PDO years in general then there is a big difference especially over the Siberian Arctic and Greenland.

 

[Rainshadow]

I agree. But his assumptions (El Niño + AO-) would provide a greater chance at such an outcome. Here is the data since 1950-51 when regular data on the AO was kept:

 

Total winters: 63

Winters with 30" or more snow: 12

19.0% had 30" or more snow

 

Winters with a DJF average AO of -1 or below: 17

Winters with 30" or more snow: 6

35.3% had 30" or more snow

 

Winters with a DJF average AO above -1: 46

Winters with 30" or more snow: 6

13.0% had 30" or more snow

 

Winters with a DJF average AO above 0: 23

Winters with 30" or more snow: 1

4.4% had 30" or more snow (1983 blizzard)

 

Winters with DJF average AO of +0.25 or above: 18

Winters with 30" or more snow: 0

0.0% had 30" or more snow:

 

Winters with a DJF average AO of -1 or below and ENSO R3.4 anomaly of +0.5°C or above: 7

Winters with 30" or more snow: 3

42.9% had 30" or more snow

 

Note: Sample sizes shrink, creating additional uncertainty.

 

In short, if one is able to reasonably estimate whether there will be strong, weak, or very little blocking, one has some idea about the probability of a very snowy winter (30" or above), but that's really about it. The majority of even strong blocking winters does not produce blockbuster 30" or above seasonal snowfall.

 

If I were trying to make an estimate, I'd refrain from providing a specific snowfall estimate or range. I might suggest an above climatalogical probability of high/low snowfall.

 

Unlike Mr. Bolaris, my early thinking favors an AO+ winter (SAI, OPI, and my own statistical probabilities). If so, prospects for 30" or greater snowfall would appear to be quite low. (I hope I'm wrong). 

Don,

 

The trends of both the statistical and dynamical models wrt to el nino have been trending away from an el nino developing (vs older outlooks from earlier this fall) and if last winter is any litmus test, they have a warm (leaning toward el nino bias). Spot on about better long term skill with temperatures than precip/snow.  

[donsutherland1]

Don,

 

The trends of both the statistical and dynamical models wrt to el nino have been trending away from an el nino developing (vs older outlooks from earlier this fall) and if last winter is any litmus test, they have a warm (leaning toward el nino bias). Spot on about better long term skill with temperatures than precip/snow.  

I agree with you, Tony.

 

Even as the ENSO regions have warmed recently, I suspect that this warming will prove similar to the late August/early September warming in that it will be temporary. I still have not seen enough evidence that the easterly trade winds are weakening on a sustained basis.

[Quakertown needs snow]

Don it's nice to see you posting in here along with the other met's.

Thanks.

[donsutherland1]

Don it's nice to see you posting in here along with the other met's.

Thanks.

Thanks Andy.

[chubbs]

I agree with you, Tony.

 

Even as the ENSO regions have warmed recently, I suspect that this warming will prove similar to the late August/early September warming in that it will be temporary. I still have not seen enough evidence that the easterly trade winds are weakening on a sustained basis.

 Agree. A strong Kelvin wave raising temos now but that will proceed to a cooling phase. Just a swag but we may end up averaging very close to 0.0 for the winter as a whole.. 

[Rainshadow]

Regardless, the return of normalcy in November is going to knock some of the pegs off of October.

[HM]

I didn't - which I guess is your point. But even though there is a predominance of +PDO in the high sea ice years and -PDO in the low sea ice years, there is still enough variability to show some sort of signal. Sadly there is a lack of years where the sea ice has reduced combined with a +PDO to formulate a true meaningful anomaly composite. 

 

However to edit, if we look at the mainly -PDO years that the low sea ice years encapsulate and compare them to the -PDO years in general then there is a big difference especially over the Siberian Arctic and Greenland.

 

VFCK7eMEaD.png

 

Thanks for the response. The best we can do is do what you are doing here. Some day, we'll have enough data to make some statistically significant relationships known. We have had a good amount of research, though, posted about the sea ice-feedback, as I'm sure you've seen already. The early ideas about how it can slow down the pattern, aiding in stagnant circulations/blocking isn't bad.

 

The thing that has me concerned about this relationship, this year, is the strong tropospheric PV in the means across Siberia. I know the sea ice/snow cover did recover somewhat this year, but this isn't why I'm bringing this point up. There aren't too many autumns that pass where we have a significant cyclonic anomaly here in the streamfunction, surrounded by poleward anticyclones.

 

Having said that, there are still signs that the diabatic circulation (BDC) is better than most of these winters that appear to generally signal "cold" ahead. But we'll see where we stand in a month. Right now, the forecasts are not promising.

[HM]

Quick update: I think there is still something about mid-December. I am not sure yet what the extent of the wintry episode will entail, but there could be a couple of storm signals showing up. There also could be something at the end of this month, which would be interesting for Thanksgiving. A lot of this depends on what the MJO-AAM-wave breaking system wants to do. Everything seems to be a little faster than my expectations in October; but, I have nothing to add really LR-wise. 

[toronto blizzard]

Quick update: I think there is still something about mid-December. I am not sure yet what the extent of the wintry episode will entail, but there could be a couple of storm signals showing up. There also could be something at the end of this month, which would be interesting for Thanksgiving. A lot of this depends on what the MJO-AAM-wave breaking system wants to do. Everything seems to be a little faster than my expectations in October; but, I have nothing to add really LR-wise. 

Is December still looking like a forecasting headache for you?