2019 ENSO

[raindancewx]

I've basically finished my winter forecast using the updated analog weighting. Snow maps are a pain in the ass, I ended up looking at something like 40 or 50 towns to try to draw the boundaries correctly.

I was hoping the ECMWF plume for Nino 3.4 would be out by now, but it isn't. If it isn't out tomorrow, will probably just post my forecast without it.

Here is the subsurface trend for the past few months. The Nino 3.4 zone is completely filled up now with warmth. It does look to me like the warmth is moving East, so the Modoki look right now will probably fade later in winter, I think that is why my analogs are showing a big mid-winter thaw nationally for a month or so.

[It's Always Sunny]

I can see a weak El Nino (or Modoki?) redeveloping in the coming months based on westerly winds pickup up a bit and ushering warmer water east.  It appears there is a downwelling Kelvin wave approaching eastern Pacific.

 

[It's Always Sunny]

18 hours ago, raindancewx said:

I've basically finished my winter forecast using the updated analog weighting. Snow maps are a pain in the ass, I ended up looking at something like 40 or 50 towns to try to draw the boundaries correctly.

I was hoping the ECMWF plume for Nino 3.4 would be out by now, but it isn't. If it isn't out tomorrow, will probably just post my forecast without it.

Here is the subsurface trend for the past few months. The Nino 3.4 zone is completely filled up now with warmth. It does look to me like the warmth is moving East, so the Modoki look right now will probably fade later in winter, I think that is why my analogs are showing a big mid-winter thaw nationally for a month or so.

How long is a Modoki cycle typically? CFS forecast through end of November weakens the Modoki like you mentioned above.

[raindancewx]

I'm not sure Modoki stuff follows a cycle like ENSO overall, especially with Nino 4 warming faster than the zones to the east over the last sixty years. I do think it is interesting that these recent El Ninos have built West to East, a lot of historical El Ninos built East to West if that makes sense, so that might be a cycle.

This is the stuff I was waiting for to finish my forecast. The trend is for Nino 3.4 to warm up somewhat from September readings, like it was last month - but Nino 4 is forecast to cool while 1.2/3 warm - so the look becomes less of a Modoki if all that verifies.

[raindancewx]

https://t.co/1yr0fAbUKK  2019-20 Winter forecast. I went with a Modoki El Nino turning to a more basin-wide event with time based on the subsurface. Snow maps, precip maps, temp maps - they are there.

https://t.co/OwLKUbp8ka?amp=1 2018-19 forecast - did some images showing verification at the end.

I'm not completely sold on this event being officially designated an El Nino, I think we only get 3-4 months at 27.0C in Nino 3.4, but its close enough honestly.

 

[raindancewx]

Definitely an idealized Modoki El Nino look for the past month, but the warm waters do seem to be pushing east into Nino 3 now. The areal extent of the cold to the east is thinning. You can also see the PDO going negative to some extent as the warm ring moves away from Alaska and the tongue that should be cold east of Japan in a positive PDO, is very warm. 1953-54 had a 23.25C DJF seasonal reading in Nino 1.2 - so I have that in my analogs to incorporate the very cold waters by Peru that will take a bit longer to warm up than Nino 3.

[40/70 Benchmark]

On 10/10/2019 at 9:22 PM, raindancewx said:

https://t.co/1yr0fAbUKK  2019-20 Winter forecast. I went with a Modoki El Nino turning to a more basin-wide event with time based on the subsurface. Snow maps, precip maps, temp maps - they are there.

https://t.co/OwLKUbp8ka?amp=1 2018-19 forecast - did some images showing verification at the end.

I'm not completely sold on this event being officially designated an El Nino, I think we only get 3-4 months at 27.0C in Nino 3.4, but its close enough honestly.

 

Great job.

Although last winter was actually pretty snowy across NNE.

[raindancewx]

The cold pool near Peru is starting to surface so Nino 1.2 and Nino 3 cooled this week. Nino 3.4/4 were down a tiny bit.

Against 1951-2010 averages in Nino 3.4, 27.0C is +0.5C, but CPC uses 26.75C, the 1985-2014 average, for Nino 3.4. Either way, Nino 3.4 for October is at borderline El Nino conditions - 27.15C or so, pending the rest of October...but lots of warmth below the surface is set to come up shortly.

               Nino1+2      Nino3        Nino34        Nino4
 Week          SST SSTA     SST SSTA     SST SSTA     SST SSTA
 18SEP2019     19.1-1.3     24.2-0.6     26.5-0.2     29.3 0.6
 25SEP2019     20.0-0.5     24.8-0.1     27.2 0.5     29.7 1.1
 02OCT2019     20.0-0.6     25.1 0.3     27.2 0.5     29.7 1.0
 09OCT2019     19.7-1.0     24.8-0.1     27.1 0.4     29.5 0.9

You can see that if we pop up to 27.15C or so in October, it's fairly close to the top 10 warmest Octobers already (27.58C is 10th since 1950).

Nino 3.4 for the past three weeks has more or less caught up to 2014-15 at this point, but 2014-15 was much warmer in Nino 3/1.2 and colder in Nino 4. At this time of year, Nino 1.2 tends to lead what the PDO will do - so the very warm Nino 1.2 in 2014 preceded the +2.07 PDO reading for Nov-Apr.

 17SEP2014     21.0 0.7     25.2 0.4     27.2 0.5     29.4 0.8
 24SEP2014     21.2 0.8     25.4 0.5     27.1 0.4     29.3 0.6
 01OCT2014     21.7 1.1     25.4 0.5     27.1 0.3     29.2 0.5
 08OCT2014     21.3 0.6     25.5 0.6     27.1 0.4     29.1 0.5

[raindancewx]

All fear the PDO? The JISAO update isn't in for September yet, but the severe cold in the NW and severe warmth in the SE are both showing up in the zones that correlate highest to the PDO for winter.

The cold ring by Alaska hasn't been there in a long time but it is trying to form.

[40/70 Benchmark]

On 10/9/2019 at 11:51 AM, Isotherm said:

A strict construction of the ENSO-oceanic analysis may reveal neutral tending toward weak nino, but inclusion of the GSDM into the analysis yields a different conclusion. The total sum of momentum additions/subtractions evinces an apparent deficit/easterly momentum dominating the tropical/sub-tropical domains, and as such, retrogressive features will be the mainstay of the short to medium term. Whether the GWO neutralizes, thereby comporting more with the oceanic appearance remains to be seen.

 

Care to elaborate? Are you implying that the ocean and atmosphere are poorly coupled like last year?

[raindancewx]

At the end of the day, you can try to forecast a season using two basic methods:

1) Find predictive variables and guess what they will do (this seems to be what most of you do). That will work most of the day, but you'll get situations like last year, where 1963 matches on most of the variables that are predictive but end up completely different by ignoring weather clues, like the super hot October 1963 compared to the very cold October 2018. My analog system, prior to incorporating weather matches had 1963 as a top three match last year...but it fell off to like 15th with weather included in Fall. 

(2018-19 was 27.40C in Nino 3.4 for DJF, 1963-64 was 27.35C in Nino 3.4 for DJF)

2) Find predictive variables. With the predictive variables find a blend that matches the weather for a long-period, with the variables and weather still heading in the "correct" direction together going forward. This is what I try to do.

I'm really looking forward to seeing the forecasts this winter - there are so many more things that can go wrong this winter compared to last winter:

1) Solar - there is a non-negligible chance of a rapid increase by mid-2020

2) PDO - it looks like it is heading more Neutral, and maybe legitimately negative, but it also may snap positive. My analogs assumed near neutral, on either side. People see the Blob and think "2013" but US weather really hasn't matched well to 2013 at all since Summer. October has been a bit closer, but I think it is a case of two trains passing each other from opposite directions. In 2013-14, the PDO was warming from years of being in the predominantly negative phase. In 2019-20, it is cooling from years of being in the predominantly positive phase.  

3) ENSO structure. Modoki El Nino since September. But Nino 1.2 is losing the cold pool below it, Nino 4 warm pool below it is moving east.

4) NAO. Pretty rare historically to have any streaks  <=-0.3 more than four months, less than 2% of all possible outcomes. October is negative. November probably will be too.

5) SOI under -10 in September is like a ~95% El Nino indicator for winter.

For what its worth, the new Jamstec (October) does have an El Nino Modoki look continuing in winter, with a cold East. The dry slot from TX to MI, with New England and the West Coast warm is pretty much what I had in my forecast.

[raindancewx]

Just for reference: Nino 3.4 is now very close to last year.

K - 273.15 = C for reference. So the 27.0C is orange, because it is 0.5C or more above the 1951-2010 October average in Nino 3.4

[raindancewx]

Nate Mantua PDO value came in today: https://oceanview.pfeg.noaa.gov/erddap/tabledap/cciea_OC_PDO.htmlTable?time,PDO

Sept 2019: +0.41

Aug 2019: +0.38

Sept 2018: +0.09

I believe October will be substantially lower than September, but we will see.

[raindancewx]

Large area of +1C to +2C heat has filled into Nino 3.4 during the last month. With the surface readings over 27.0C in October...that's enough for Nino 3.4 to stay warm for a while. The cold in Nino 3 (90W-150W) is just about gone now too, although Nino 1.2 is still going to take a while to warm (maybe til December).

[The_Global_Warmer]

You would think we would be setting up for a major EL NINO heading into Spring next year.

 

But the sub surface while leaning warm is not primed for strong Nino. 

 

[raindancewx]

It's not exact, but in some ways we're about a month behind the development of the El Nino last year. The cold by South America was mostly defeated by this point last year.

But...it was still there in September 2018.

The Oct 2019 pattern by South America looks a bit like the blend of Sept/Oct 2018.

[raindancewx]

Bye bye Modoki? Nino 1.2 has warmed up above 20.0C - so the PDO likely will remain positive, at least slightly for Nov-Apr. 
El Nino development is around a month slower than last year so far. But October Nino 3.4 reading is up to about 27.3C now.
That's El Nino level, even using the 26.75C CPC uses for October as Neutral. Subsurface warmth is up again too.
It looks like the subsurface heat for 100-180W could be around +0.6 or +0.7 - that is similar to October 2004 & October 2014.
             
                 Nino1+2      Nino3        Nino34        Nino4
 Week          SST SSTA     SST SSTA     SST SSTA     SST SSTA
 25SEP2019     20.0-0.5     24.8-0.1     27.2 0.5     29.7 1.1
 02OCT2019     20.0-0.6     25.1 0.3     27.2 0.5     29.7 1.0
 09OCT2019     19.7-1.0     24.8-0.1     27.1 0.4     29.5 0.9
 16OCT2019     20.6-0.2     25.3 0.4     27.5 0.8     29.7 1.1

 26SEP2018     20.2-0.3     25.5 0.6     27.3 0.6     29.3 0.6
 03OCT2018     21.3 0.7     25.6 0.7     27.4 0.7     29.5 0.8
 10OCT2018     21.1 0.4     25.6 0.7     27.3 0.6     29.5 0.9
 17OCT2018     21.1 0.3     25.9 1.0     27.6 0.9     29.6 0.9

[40/70 Benchmark]

Thanks for the great info, raindance, as always.

2004 and 2014 are two modokis I do not mind mirroring.

[raindancewx]

2004 is probably a good analog to November. If I had done Fall analogs, 1994, 2004, 2012, 2017, 2018 as a blend would have been pretty good. But I don't really see it lasting into winter. All the components are there - MJO matches 1994, 2004, 2018. The cool east is there with 2017 included. The temperature profile in the US looks like where Oct 2019 is going. The PDO is neutral. The IOD is positive. The coolness east of South America is there, with the gulf stream very warm off the East Coast. It's relatively low solar too.

2014 doesn't exactly look like 2019 in October...despite somewhat similar MJO progression...it makes sense since 2014 was a basin-wide El Nino in October. Only off by 10 or 20 degrees in the NW...but you know New England is destined to see snowmaggedon again. 2004 isn't as different in October as 2019, but it's still kind of north/south backwards, just a 2014 v. 2019 is east/west backwards.

[40/70 Benchmark]

No one is calling for a snowmageddon in New England, but I don't see any awful signs is the point.

[It's Always Sunny]

@raindancewx I tried clicking the link for your seasonal forecast but it's not loading? I tried Chrome & IE. Is it subscription based only?

[Typhoon Tip]

how about "notsnowmeggetin ' " 

does anyone ever have the plumbs to make that call ? 

[raindancewx]

23 hours ago, It's Always Sunny said:

@raindancewx I tried clicking the link for your seasonal forecast but it's not loading? I tried Chrome & IE. Is it subscription based only?

I couldn't get it to work at my office either. It is a link to Scribd, so maybe it is blocked? I think some have been able to see it. It is free. But I may do some subscription stuff at some point. I have a big collection of strong predictors for each region of the US now, and I've done statistical work on how they should be weighted for different seasons, and to account for the Earth warming somewhat, as well as solar stuff. For instance, in the Southwest, solar activity is a meaningful precipitation variable for Spring/Fall, but not really for winter. But in winter it does correlate for temperatures, when filtered by ENSO, and it gets stronger at certain elevations for whatever reason.

https://www.scribd.com/document/429712619/Winter-2019-20-Outlook

I want to be clear here: There are some El Ninos when I would go huge for snowfall in New England and the Northeast generally. I don't mean 20-50% above average, I mean double average generally, up to x3 or x4 locally - but they're rare. The blend you want is high solar, El Nino, but El Nino Modoki. 2014, 2002, 1991, 1977, 1969, 1968, 1957 all fit some extent.

For what it's worth...1951-52 is an El Nino and remarkably close to October 2019 so far. It's kind of a weird borderline El Nino like this will be, which is interesting. I didn't use it as analog though. 1994 is actually pretty close in the West, and 2018 isn't too dissimilar either in October, its just the heat/cold are nudged out of place compared to 2019. Some of the El Ninos in the 1930s and 1940s aren't terrible matches to 2019 either for what its worth. If you're in the East, you do not want the 1951-52 winter if you like snow and cold.

[40/70 Benchmark]

2004 and 1976 were pretty epic in SNE, too.

I would agree that this isn't a clear-cut blockbuster season, though.

[Hoosier]

3 hours ago, raindancewx said:

I couldn't get it to work at my office either. It is a link to Scribd, so maybe it is blocked? I think some have been able to see it. It is free. But I may do some subscription stuff at some point. I have a big collection of strong predictors for each region of the US now, and I've done statistical work on how they should be weighted for different seasons, and to account for the Earth warming somewhat, as well as solar stuff. For instance, in the Southwest, solar activity is a meaningful precipitation variable for Spring/Fall, but not really for winter. But in winter it does correlate for temperatures, when filtered by ENSO, and it gets stronger at certain elevations for whatever reason.

https://www.scribd.com/document/429712619/Winter-2019-20-Outlook

I want to be clear here: There are some El Ninos when I would go huge for snowfall in New England and the Northeast generally. I don't mean 20-50% above average, I mean double average generally, up to x3 or x4 locally - but they're rare. The blend you want is high solar, El Nino, but El Nino Modoki. 2014, 2002, 1991, 1977, 1969, 1968, 1957 all fit some extent.

For what it's worth...1951-52 is an El Nino and remarkably close to October 2019 so far. It's kind of a weird borderline El Nino like this will be, which is interesting. I didn't use it as analog though. 1994 is actually pretty close in the West, and 2018 isn't too dissimilar either in October, its just the heat/cold are nudged out of place compared to 2019. Some of the El Ninos in the 1930s and 1940s aren't terrible matches to 2019 either for what its worth. If you're in the East, you do not want the 1951-52 winter if you like snow and cold.

Interestingly, Nov 1951 produced some significant snows unusually far south.  For example, Chicago had 14" with virtually all of it in the first week of the month.  Next week looks to have some significant snow potential in the Plains/Midwest.

[raindancewx]

My analogs had some incredible cold pushes in the West and Plains in November, despite relatively bland departures overall. A big -NAO is correlated to a warm West in November generally but you can get big snow events in the Rockies despite an overall warm November.

Conceptually, the "anti-logs" to the features this year, a cold Atlantic, cold North Pacific, weak La Nina, after a weak La Nina, both in high solar, with an active Monsoon in the SW ahead of winter...is something like 1967 or 1971. If you look at October 2019 and then October 1967 and October 1971, they are both opposite of 2019 in their own way.

In 1967, the core heat/cold is East/West backwards v. 2019 in October, even though it is roughly at the right latitudes. Winter 1967-68 "flipped" for being east-west backwards actually looks fairly close to what I forecast for winter.

In 1971, the cold core is in the SW, and the heat core is in the NE. In 2019, the cold core is in the NW, and the heat core is in the SE. So it is North/South opposite. Winter 1971-72 flipped "North-South backwards" actually also looks fairly close to what I forecast for winter.

[40/70 Benchmark]

I def. agree with the translation eastward of the developing el nino...much like last year. That is that main difference between this season and 2014..el nino is developing a bit later and is working from west to east, whereas that year it was east to west. Otherwise, 2014 is a good ENSO analog.

[LibertyBell]

5 hours ago, 40/70 Benchmark said:

I def. agree with the translation eastward of the developing el nino...much like last year. That is that main difference between this season and 2014..el nino is developing a bit later and is working from west to east, whereas that year it was east to west. Otherwise, 2014 is a good ENSO analog.

I was thinking backloaded winter too, with the "back load" starting later- perhaps in the first week of February?

 

[40/70 Benchmark]

3 hours ago, LibertyBell said:

I was thinking backloaded winter too, with the "back load" starting later- perhaps in the first week of February?

 

Don't completely agree...

[LibertyBell]

58 minutes ago, 40/70 Benchmark said:

Don't completely agree...

Well have a look at the Euro LR maps, Ray, they look like they're forecasting January to be below normal snowfall.  Maybe normal snowfall for some areas.  February is when they're depicting well above snowfall.

https://bennollweather.com/ecmwf

https://www.bennollweather.com/ukmet

 

UKMet has a better snowfall distribution.