raindancewx

Here is where winter stands ~1/9 through winter. Suspect the reds will hold on in Montana and Arizona through Dec 31, but the rest of the country is up for grabs if we can get the MJO coherently into phases 7/8/1/2/3 by the end of the month. The snow forecast in the interior NW over the next 10 days or so makes me think that area will stay cold.

Going to be very interesting seeing how the high temperature anomaly for December finishes in Albuquerque - January is heavily favored to have the coldest anomaly here in low-solar La Ninas, and to a lesser extent in low-solar years generally. Signal is less strong for whether Dec/Feb ends up warmest. I went -1.5, +0.0, +1.3 for Dec, Jan, Feb, between a warm Nov/Mar because assumed it would snow in December, but the raw analogs had -0.8, -1.2, +0.1 for the three months.

Monterrey, MX supposedly this morning.

Could've sworn I said somewhere I thought there was a shot at a blue norther interacting with the STJ, that's kind of what has happened, but can't find it. If the MJO stays active, and I think its magnitude is correlated to some extent to the QBO and Solar Conditions - this setup should return later in the winter, but probably further to the north, which could be a good set up for Amarillo, Dallas and Eastern NM. Neat to see #houstonsnow trending on Twitter. Lots of people saying its their first time seeing snow falling ever in San Antonio and Houston.

I love watching people deep in TX/MX and the South react to rare snow. It's neat to see. The 2008-09 winter was an analog that I used for my winter outlook, and it snowed in Houston & New Orleans - so not too surprised to see it. It was a low solar La Nina, after a La Nina, two years after an El Nino, and with impressive Gulf of Mexico hurricanes. https://twitter.com/deweythoms/status/938942069116194816 Accumulating snow in San Antonio https://twitter.com/mattbonner_15/status/938934812106162176 https://twitter.com/damaya14love/status/938932177663922176 Snow in Austin https://twitter.com/KHOULauren/status/938953217521672195 Snow in College Station, TX

Its amazing seeing snow forecast at short ranges SW of Brownsville TX in MX, and also by New Orleans, Houston and other places that almost never see it.

Looks like November, which was hyped to be pretty cold, ended up only slightly cold in the East, with a look somewhat similar to 2007, 2012, 1949, and maybe 1995 distant 4th. It certainly looks nothing like 2009, 2010, 2011, 2013, 2014, 2015, 2016 or other recent winters with memorable features for the East. The blend of the 2007/2012 gives some cold in the Lakes/NE and a lot of heat in the SW for November which is the right idea even if the core of the cold/heat is off a bit. A blend of 1995/1996/1999 isn't bad either, two cold years in the Lakes/NE v. one super torch. Going to be hard for December to be too cold nationally, with 20% of the month registering incredible warmth. The cool area to date (NV) is where the Canadian had it cold to date. The reds are fading some in the West generally, with the high here forecast at 37F or so on 12/7, which is ~11F below normal for the date. The current look of the map for December is close to what my analogs had for January, but the MJO progression has been different from the years I used, so its not super surprising. The Canadian look for Dec (Nov 30) had the cold centered on NV, with NM/CO/WY on the edge of it/near average and everyone else quite warm.

This setup is early days of "pattern b" to me. One of the three patterns my analogs had for the winter. The precip by El Paso and up the entire East Coast is consistent with it.

King NAM has snow for much of TX, NM and NE MX.

Not sure how it will hold up, but my analogs had the cold East/hot West look in Nov, and then a gradual reversal to what the Canadian shows in December, probably Dec 15-25. I would offer that with the complete lack of rain/snow in Albuquerque since Oct 5, and the record for no-precip being 109 days since 1892, any rain/snow that makes its way deep in the SW will be the indicator of the change. Also, November in La Nina years is a pretty poor indicator for winter anomalies. I know people like 1995-96 in the NE, but Nov 2007 is close too, the cold in the East in 1995 was much more intense than this year, but 1995+2007 isn't bad for Nov.

It has snowed some in the northern mountains, its the valleys and central areas that are lacking. We're in a top-30 dry stretch for consecutive days without precipitation in Albuquerque back to 1892. The record is 109 days, currently at around 60. Would assume the streak ends way before 109 days though.

It snowed a lot in Mexico in the 1960s...and the best objective match to Nino 1.2, Nino 3.4, and Nino 4 SSTs in November was a blend of 1966 (x3) & 2007 (x2). So I'm on board with Mexico getting some real, accumulating snow this week.

King NAM and the latest GFS each have some snow in West TX and SE NM. Hoping it trends North.

Didn't really "design" my analogs for November, but the years I picked for winter (in Oct) did pretty well for the month. Anomalies will be closer once the maps become available only for high temperature departures, which is all I care about, lows are just about always way warmer than older years now.

Canadian model has trended somewhat colder (to only a bit above average) and much wetter for this area of the US, as the Jamstec did last month, and as I forecast(!).

If the MJO really does get to phase five at coherent magnitude, I have little doubt we have at least some kind of cold spell. 18z GFS and most recent Euro were back to some kind of precipitation event for NM, but still not looking great overall.

If the Modoki value stays around 0 we'll probably stay fairly dry and warm here - if Nino 1.2 warms before Nino 3.4, we could get wet quick. If Nino 3.4 warms before Nino 1.2, would think we cool, but stay dry. I had the Modoki value at -0.2 for winter, that doesn't look great right now with Nino 1.2 so cold, but would shift the cold East into the Midwest from where I had it by NV/UT/ID.

http://weather2020.com/2017/11/27/the-2017-2018-winter-forecast/ I like some of the ideas Gary has for long-range stuff, but to me he gives too much power to Oct-Nov. If Oct-Nov just repeated over and over again without much more than a shift north/south, we'd get no rain where I live for a year? That doesn't happen (we had 0.04" in Oct-Nov, lowest 12-month total is maybe 3.3"?). If you included Sept 26-30, when we had 2.2", then we'd look wet, since we had 2.2", but he doesn't. To me the patterns reset in July when the Southern Hemisphere is in winter, that is when the changes in Nino 1.2, south of the equator, start to bleed into our hemisphere to set up the pattern. I think the MJO activity this winter is going to be on/off this winter but at short intervals, so at some point (maybe Dec 5 as the models hint) Dec 5 is in the pattern for winter via the late Sept set up.

GFS - 8 day snowfall on the left. Euro precipitation on the right (as snow) through 10 days. Both have 4-6" for even ABQ in the fantasy range.

I've never looked into this, but does TX tend to have dry winters after big hurricane years? Certainly seems like it in some of the analogs I liked. My analogs for winter v. November 2017 so far - Correct pattern, not great on anomalies.

My winter analogs, applied to November, produced a near perfect match of temperature anomalies for the last week. Don't think the month will end up being quite this close unfortunately. South started too warm, NW started too cool. Sometimes I think I should put in some kind of timing variable for analogs, but not sure how.

Been doing research trying to predict the PDO. Seems like the March-August PDO values (Mantua) are either "eaten away" or "reinforced" by whatever the temperature anomaly is in Nino 1.2 during October. In other words, if its a positive PDO in March-August, but Nino 1.2 is cold in October it will trend to lower values in winter. If it is a positive PDO March-August, and a warm Nino 1.2 in October, it stays positive or becomes even more positive and so on. Well anyway, we had a positive PDO in March-August 2017 (+0.62), but a cold Nino 1.2 in October - and so the general idea on my end was that Nov-Apr would trend down from Mar-Aug levels. October PDO value just came in --> +0.05 --> lowest value since December 2013. Depending on the data-set you use, the blend of Mar-Aug 2017 and Nino 1.2 implies +0.1 for the PDO in Nov-Apr, +/- 0.4, at around 90% certainty going by 1994-2016 (excluding 1997, since Nino 1.2 has never been that nearly warm since 1930)

The JAMSTEC has switched to a slightly cold winter for much of the SE / SC US. Kind of went from what I expect (cold interior West) to a blend of what cerakoter had and what I had. Very different from what it forecast last November.

It won't be of much use for this winter, but I've been working on a "Modoki" calculator for winters - will be useful if we go into an El Nino. Works fairly well using Nino 1.2, Nino 3.4, Nino 4 SSTs (raw) in October, with a recognition that each Box, Box A, B, C, must be calculated separately for the Modoki number to work correctly. The weak La Ninas and weak El Ninos seem to have their cold directly controlled by the structure of the Modoki signature, and with Box C warming quickly over time relative to the other boxes, certain types of events are becoming more likely.

My winter analogs weren't really "designed" for November, but they had everyone east of a line from Biloxi to Billings cold, and everyone west of that line warm. So far the cold has been more centered over Montana, but it does look like it will drain East for a while now.