raindancewx

The oceans aren't a great match to 1978, but the Atlantic was closer than I expected it to be. 1983 is fairly close, but off in certain spots. I like a blend of 1937, 1943, 1959, and 1983 for the oceans as an ~80% match. It's a bit off in the Western Atlantic/Western Pacific. I formalized my "June to November" observed highs v. forecast thing - for the last 30 years, it shows a small amount of skill for forecasting seasonal temps here. It's something like 80% confidence at +/-3F for it's DJF estimate (range is 12F historically, 44F-56F, so it cuts out maybe half the uncertainty?), and it's still ~67% at 2.3F. It's about 15% better than guessing the average (49.5-49.6F) each year. Anyway, it says ABQ has a mean high of 48.6F - and thus an 80% chance we are within 3F of that. Against the 1931-32 to 2015-16 winter time frame, that's -4F to +2F - a full 2.3F colder than last year, when the hindcast came up with 50.9F +/-3F. The Spring has either less variability or this is a better tool for Spring, as it's +/-2.25F for MAM 80% of the time, and +/-1.8F 67% of the time. Warm Spring is slightly favored in my area, but likely near average. I keep harping on 1978 because it is a very strong match for my last six months of temperatures here - but in my area it wasn't a particularly cold winter. I'm somewhat partial to lumping winters after two El Ninos in a row together - it takes longer to cool the oceans down than in a normal El Nino to Neutral/La Nina transition. So I look pretty hard at the winters starting 1941, 1942, 1959, 1970, 1978, 1988. Just from that list, 1942, 1959, 1978, and 1988 are all somewhat similar to now, at least for my observable weather.

Do any of you like 1978-79 as an analog? It was a borderline La Nina (-0.4C) in Fall of 1978, but it weakened to Neutral in winter. It followed two El Nino winters (76-77, 77-78), and occurred after a reversal in the long standing sign of the PDO (1976/2013 - although the more recent flip seems to be a spike more than a 30-year flip). I ask because the La Nina seems to be weakening, we're following two El Ninos...and the cold experienced in the US in January 1979 (-8F to -20F over most of the US) has been occurring over Asia for the past two months. January 1979 was stupid cold in the middle of the country - including TX, OK, etc.

Some flurries around Albuquerque tonight. Don't think it stuck anywhere - very light, above freezing. Always amazing to watch the dry air do it's thing, went from 37F with a 27F dew point to 35F with a 29F dew point and snow in about two minutes. Fell to 33F with a 30F dew point before it stopped snowing.

I was hindcasting my "objective" match testing the other day by mean highs, and although it doesn't work nationally, it does seem to show some skill from roughly TX to AZ to OK to Southern UT/CO/NV, and then quite a bit of skill the closer you get to ABQ. It doesn't really show any skill for precip, just temperatures. What I wanted to show though, was that last year, the temperature profile behaved more like a blend of La Nina years from June-November, and then remained La Nina-ish/warm in Winter and to a lesser extent Spring. The temperature blend this year, despite being nominally in a La Nina, is much more of a weak El Nino / warm Neutral signal, which favors a much colder winter, maybe ~2.5-3.5F colder than last year. The six closest mean high transitions last year from June to November favored a winter (Dec-Feb) mean high of ~50.9F in Albuquerque. It was 51.4F. The average is 49.5F. So the method identified slightly warmer than normal correctly. This year, the method says 48.6F - which is colder than normal, and 2.4F colder than last year.

Panel A is the best match to my mean highs from June-Nov carried forward to Dec-Feb Panel B is the best match from June-Sept mean highs, the monsoon, ENSO, years since ABQ had >9.6" snow, and the monsoon+ENSO Panel C is the best match for my winter ocean expectations - Cold Neutral to Weak La Nina, Modoki, Neutral PDO, Warm AMO, after two El Ninos, Dry Monsoon Panel D is the best match for borderline La Ninas only: -0.4C to -0.6C in DJF So far, A has been pretty good for Nov, with B good for Oct. My historical testing of these methods show "C" should become most accurate later in the winter, with "B" fading after December to irrelevance. D should be relatively decent. But I lean towards a split of "A" & "C" for the winter, plus 1-2F in the SE. Essentially, A early, then C late. Anyway, best bet for the West: WA/OR: slightly cooler than normal (-0F to -2F) ID/WY/MT: pretty cold (-2F to -4F) CO/WY: cool (-1F to -3F) CA: normal (-1F to +1F) AZ/NM/NV: normal (-2F to +2F)

Looking back at non-El Nino years that were relatively wet in Albuquerque during Sept-Nov produces an interesting list of years: 1935, 1938, 1978, 1983, 1988, 1992, 1998. Barring more rain, or our first snow this month, those years had 2.4-3.2 inches of precipitation from Sept-Nov. We're at 2.8 inches of precip in Sept-Nov 2016. I consider 1938, 1983, 1988, 1998 to be La Ninas. 1992 & 1935 are sort of warm Neutrals that wanted to be El Ninos, with 1978 following a double El Nino as a Neutral year. This is probably the wettest November NM has ever had in a La Nina, so blending in Neutral years seems appropriate. 1978, 1983, 1988, 1992, 1998 - these years all followed an El Nino. Pretty cold winter in the US if you blend those seven years together in the winter. Most areas 1-3F below normal, 3-5F below in Montana and the Upper Midwest, +/-1F in the SE. Very wet in much of the west & the east, drier from El Paso to Detroit.

This is an interesting forecast idea - and consistent with what I put out in October: - Warm early - Colder late - Lots of variability for where cold/warmth is in any given moment https://riskpulse.com/blog/2016-2017-winter-weather-outlook/

Lots of lightning, heavy rain, mist and hail here tonight in the Land of Enchantment. Snow above ~7,000 to ~7,500 feet. Wettest November in Albuquerque since 2004 with over a week to go.

Now that its getting relatively cold - two frosts here in ABQ) - I dug out & updated my "gambler's snow tables" for Albuquerque. It's a good way later in the season to see if the computer models are spitting out bull. The table reminded me how awful most La Ninas are in the Southwest for snow. I'm 5,300 feet above sea level, but the odds of getting even 3.5" snow in a MONTH, not even a storm, a MONTH, are only a coin-flip in a La Nina based on 1931-32 to 2015-16. One of these years we're going to get actual notable precipitation in March again and the items I've marked overdue should begin to fall more readily. We haven't had a March within even 20% of the long-term average (0.5") since 2007, so it would be nice if it happened this year. One thing is for sure, given the bubkis precipitation we get last March, even as it snowed in Guadalajara - check out Nieves en Guadalajara on twitter - we can't get less precipitation in March than last year.

The JAMSTEC (Japan's model) has trended much wetter for NM/CO/TX for the winter. New Mexico went from 0.1-0.6mm below normal/day for 90 days (0.35-2.1 inches of precip below normal) to w/in 0.1mm/day of average for 90 days, i.e. w/in 0.35" of normal. Pretty big shift towards moisture for us. The West Coast looks a lot drier than before too, as does the East Coast. Bit cooler too, from "very warm" (+3F) to "slightly warm (+1F).

If it makes any sense, I think the pattern of 1983-84 is a decent match, but the warmer Earth/less ice against 1983-84 is part of why I have the winter fairly backloaded away from the east coast. I had NM/AZ and much of the interior west/mountainous areas warm Oct-Dec. Doesn't really look wrong. I do think TX starts to cool off a week or two before we do though. The cold coming off Asia should eventually cool the Pacific enough to allow the Arctic ice to build up pretty rapidly - especially since I think this winter has real blocking...but it is transient and no where near permanent.

The PDO went up in October (JISAO) to 0.56 It took me a while, but Dec-Feb 1937-38, 1943-44, 1959-60, 1983-84 work well as a good match for the oceanic pattern right now. Those years had a PDO value in Nov-Apr of 0.58, an ONI value of -0.28 in DJF, and an AMO value of 0.146 from Nov-Apr. Central/Modoki La Nina, weak, cool pool off Japan & Russia, West Coast warmth wrapping west south of Japan cold pool, stripe of warmth along US east coast, cooler in eastern Atlantic. Oceans fairly cold south of the equator. It's not a perfect match, but eyeballing it, I'd say it's 80-85%?

Looks to me like the PDO is heading positive again, with the AMO weakening a bit. Both good for New Mexico & the SW. The La Nina is still pretty clearly a Modoki, and weak. Not sure how long it lasts either, I think it may peak Oct-Dec, or Nov-Jan.

This will be the latest it's ever dropped to 40F or lower in ABQ too. Can't speak for TX, but warm Octobers, especially if October is nearly as warm as Sept, tend to favor big snow late (Jan-Apr) and cold/snowy February in particular. This actually seems to be known, as it is a saying. Haven't tested the other folklore though. http://www.stormfax.com/wxfolk.htm We've had (already) a wet November here - which is pretty rare in La Nina / Neutral years, but it does tend to favor near normal snow overall.

Albuquerque (airport) got 0.66" precip Nov 1-5. That's already well above our monthly average of 0.46" precip. I'm a bit worried about my call for a warm November now. Every inch of precipitation in November lowers the mean high in November by 3.9F historically, so we're now favored to be a degree or so below normal for the month - and that's assuming it doesn't rain or snow again for the rest of the month. The most similar years since 1931 with this much precip in early November here are 1933, 1946, 1957, 1986, 1990...and those years averaged 0.97" precipitation. Snowpack also recovered a fair bit with the recent moisture.

The CPC had weak La Nina conditions for the second tri-monthly period in a row. July-Sept has been updated to -0.6C (from -0.5C) and Aug-Oct is listed as -0.7C. I'm expecting the La Nina to peak in Sept-Nov, hold fairly steady in Oct-Dec, and then weaken rather quickly. May make a good run at -1.0C though (could peak at -0.9C or so). Should be a "stronger" La Nina than the 2014-2015 El Nino was in terms of deviations from 0.0C anomalies. Part of me thinks we go into a stronger La Nina next year after a brief warm up, but there isn't a whole lot of evidence for that at the moment. It's definitely a Modoki - Nino 1.2 is currently around +0.7C by tropical tidbits, and the western tropical Pacific is warm too.

February 1964 was incredibly cold in the Southwest too, after being very warm in October. We're finishing our 4th warmest October (mean highs) and 2nd warmest October (mean temps) in ABQ. For whatever reason, February seems to favored for snow in the Southwest if the warmth of September lingers into October, or just generally if October is very warm. If nothing else, in ABQ our ten warmest Octobers don't seem to indicate warm winters at any especially high rate - it's 2/10 that are at least +2F for mean highs, against 18/75 for winters following an October that is near normal or cold. Not statistically significant. I still like warm Oct-Dec, cold in either Jan or Feb, and a decent snowstorm in late Jan or early Feb for my area, and probably for much of TX & AZ. I think for ABQ mean highs end up very close to normal (49.5F) for Dec-Feb, even after a warm start in Dec.

I had the southern 1/3 of the US wet in Nov, and the East Coast wet in Nov, so this is promising to me:

For what it's worth, Albuquerque (5,350 feet at the airport) hasn't had a freeze yet either. First freeze date (<=32F) varies a lot here, but is always in Oct or Nov, and usually right around Halloween here. Our coldest temperature this fall is still from Sept - got down to 42F...on Sept 24th...a month ago (!). First freeze will hopefully come by Nov 10. For mean highs since 1931 in Albuquerque, this is likely to be the fourth hottest October - behind 1950, 1952, and 1979. The period from July 1947 to June 1957 was incredibly dry here, with notoriously warm late summers and early falls, so the 50s being over-represented makes sense to me, and the late 70s had very hot/dry Julys too. For whatever reason though, a warm October (for mean highs) is actually pretty decent for winter snowfall here - but it does take a while to cool off, not til Jan/Feb. These are the warmest Octobers in Albuquerque, followed by snow. Also worth noting, when the mean high drops off by a small amount from Sept to Oct, (Sept mean high-Oct mean high<=9.6F) Feb snow is favored here too. The 2016 Sept to Oct drop off will be 83.2F to ~76F. Snow of over 3" in February isn't super common here, so the difference between heavy snow following a very warm Oct, when highs avg over 75F, and all other Februaries is statistically significant, 5 out of 10 against 14 out of 75, generates a P value under 0.05.

I think the warmth is definitely part of the winter pattern, my idea has been that the West would be quite warm in Nov, and especially Dec, with the east quite warm in January and to some extent into February. My issue with permanent warmth is the Summer - there were cold spots, and they were widespread at times and they moved a lot. I think we see a lot of movement in where the cold is this winter too. September also saw incredible heat in the East, with the West relatively cool, which I think bodes well for March or April.

There is a long post associated with this, but Weather Trends 360 says 1960-61, 1966-67, 1983-84, 2010-2011, 2014-2015 is fairly similar as a blend to their official forecast for Nov-Apr. Does favor a fairly cold winter for NM & CO, with much of the rest of the West warm/near average. Those years are pretty cold overall. Similar idea to what I had early, East Coast cold Nov, it backs west Dec, reaches SW for January. I think the east is warmer and the cold is further west in Jan-Feb though. March I think will be a bit more crazy looking, and then April I'm pretty different for that, but it's early. Overall, I like the idea they have, just think the cold is centered further west.

Looks like Sept 2016 is a pretty good match for Sept 1931 + Sept 2005. Usually a lot harder to produce a decent match than from just using two years. Those of you in TX shouldn't look at the winter though...

This thread needs a bump. Starting to get a bit psyched for the winter - had highs in the low 60s the other day. August & October is already wetter than last year, which favors bigger late season snows here. In Albuquerque we're up to 1.58" for Aug & Oct, and nearly at the monthly average for October rainfall already. If we get another 1.1" precipitation in October, March starts to look good for us in the SW.

Here is my Monsoon season (June 16-Sept 30) review for Albuquerque, with some early thoughts on winter: Monsoon Highlights for Albuquerque: June & July: Hot Enough To Make the Devil Cry: - Mean Highs in June & July were 3-4F above normal - 32 days with highs of 95F or hotter in June & July (1931-2015 normal: 18) - Extremely dry from June 16-July 28: 0.30" rain...against 1.75" normal The Wet Week: - July 29 - Aug 4: 1.34" rain (normal: 0.42") August: - Not that wet in Albuquerque (0.86") but wet almost everywhere else in NM - Cold: Lowest temperature of 55F was w/in 5F of August record low (1931-2015) - Cold: Lowest mean lows since 1992 - Cold: Mean high of 85.8F well below 1931-2015 mean of 89.3F (-3.5F) September: - Moist: 1.04" precipitation in Albuquerque (exactly 1931-2015 average) - Mild: Mean high of 83.2F (Historical Sept mean high = 82.6F) - Warm: After not reaching 95F in Aug, it got to >=90F in ABQ two times - Vs. Last Sept: Much colder - it never dropped below 55F in Sept 2015, was >=90F nine times (!) - Near Record Cold: Lowest temperature of 42F was within 7F of Sept record low (1931-2015) Overall June 16-Sept 30 2016: - 3.09" rain (4.31" = 1931-2015 normal) - Biggest mean high drop off ever from July to August (1892-2015), 95.6F to 85.8F - June 16-July 28: 17.1% of normal rain - July 29-Aug 4: 3.2x normal rain - Aug 5-Sept 30: 1.47" rain (2.42" normal) - Near reverse of June-Sept 2015: Hot early, cool late. Early Winter Thoughts: - Oct/Nov snow unlikely - a warm/very wet Oct would favor a big Feb-Mar for snow though. Nov may be wet. - December likely warm & dry (it tends to behave similarly to June 16-July 15) - January likely cold and wet (tends to be similar to July 16-Aug 15) - February likely near average (mild, moist). Snowy if Oct mean highs drop <=9.6F from Sept mean highs - March likely mild & dry (moisture favored when Aug & Oct see lots of rain, Aug was dry in ABQ) - April likely cold & moist (one notable snow (>=2") more likely than usual) - May likely mild & dry (we've had three cold Mays in a row) - Pattern of frequent light snow, not one or two big storms - Outside chance pattern mimicks 1958/1959 and we get a huge snowstorm in December, in a very warm Dec. - Years of most similar Summer high temperatures (June-Sept) favor a lot of snow in Feb. - If winter highs mimicked Summer, would see near record heat early (Dec) follwed by major cold snap (Jan/Feb). - If lows mimicked Summer, would see near/record lows late in season Jan 15-Mar 15. - Pattern looks disfavorable to snow during the day, but should be efficient for snow at night. If significant snow is to fall, best bet is Jan 15-Apr 15. Oct-May season likely below average though. Will release a cold season (Oct-May) forecast soon. Spring likely dry.

I wrote to Nate Mantua earlier this week, and he said the PDO value for August was 0.52, even though it hasn't updated on here - http://research.jisao.washington.edu/pdo/PDO.latest I'm going with a neutral (slightly positive) PDO for the winter. It does look pretty mixed up at the moment with the cold ring along the West coast that is consistent with a -PDO half there, but half not, and the patches of warm/cool anomalies not truly positive everywhere.