raindancewx

I've been talking about 1993-94 as an anti-log because it specifically fixes some of the issues with 1972-73. It's coldest where the models have the Atlantic warmest, and warmest where the NE Pacific is forecast coldest. So subtracting it out from the analogs fixes some of the issues - The N Atlantic (10N, 30W) and N Pacific around 140W are fixed specifically with 1993-94 included. I was doing more testing today. Actual US weather matches pretty well to a 1972, 1991, 1997, 2009, 2019 blend - if you recognize that the 1972 and 2023 El Ninos started at different points in Spring. So Feb-Mar works better with anti-1972. But you can more or less replicate the spatial placement of the heat / cold each month with that blend. January is a poor match though. You do generally also have to warm up temps by 1F or so since the analogs are centered on 1998. The QBO/Solar stuff works fairly well those five. 1972, 1991, 2009, 2019 are all neutral to very negative QBO winters, with 1997 barely positive. PDO is basically neutral. Solar is very high in 1991 and fairly high in 1972 which offset the very low 2009 and 2019. The issue with this winter is the temperature magnitudes. I'm relatively confident in some kind of north / south split for US temperature anomalies. The matches I have all look like the opposites of the +PDO / La Nina winters, which makes sense to me. I could see someone in the Northern US finish 5-10F above normal. But I also wouldn't be shocked at +1 to +3F for both cold / warm departures.

Summer highs locally filtered by ENSO third (La/N/El) tend to predict snow totals pretty well locally. I got 15 inches for Albuquerque with a slightly cool June, blazing July, and average August. If that's about right, it'd be a decent season for snow in the NE US. Snow totals are pretty strongly negatively correlated between ABQ and Philly in El Nino. I saw a video on Youtube where Bastardi said his tentative winter analogs are 1957-58, 1965-66, 1994-95, 2002-03, 2009-10. I generally think the pool of ~28C ish El Ninos are right. He's at the right strength level with that group. It's just not the PDO/Modoki setup I expect.

Here is how we look in July 2023 v. some other major events. I generally look for "red" in Nino 1.2 as a requirement for a good visual match. July sort of looks like a 1972, 1991, 2009 mix to me. Sprinkled in 1997 and -1993 as well to fix some stuff - Winter is about right conceptually - very warm Atlantic. 28C ish El Nino. -PDO. Some cold water relatively west of Australia and by India.

Some of my conceptual stuff is starting to firm up for winter. May as well share. - December in the Southwest tends to be cold following cold/wet July. We don't have that this year. - The ( +ENSO / -PDO ) subtracting out (-ENSO / +PDO) years have a strong signature for a pretty severe cold period in the Fall, likely after mid-October. - Record upper level moisture (Tonga), with a strong subtropical jet should lead to a very wet winter almost everywhere in the US. But for most places, I don't see many particularly cold storms if the PDO remains negative, especially with the AMO/Atlantic so warm. - There should eventually be decent snow pack though in cold areas, and when storms can pull in those air-masses while vacuuming up all the extra moisture around, it's not hard to imagine 5-6 explosively powerful storms in Fall and March. - I'm not expecting major temperature deviations this year for the US in winter. November and March are different. I expect large portions of the US to see top 10 type wet winters though. - I think the QBO is kind of dumb as a factor, but the -QBO El Ninos are interesting (these are yours +5 to +20 starts in January, that were negative by winter) 1958-59, 1972-73, 1976-77, 1986-87, 1991-92, 2009-10, 2014-15, 2019-20 Super cold Fall does show up with the QBO though. The US is actually pretty cold for once in Nov-Jan, but the cold is more of a Mexico thing for Jan-Mar. - El Ninos following major (VEI4+ tropical) volcanic eruptions (0-3 years after): 1963, 1965, 1982, 1991, 1994 - Warm AMO / Summer Heat Wave El Ninos / El Nino-ish years: 1951, 1953, 1957, 1994, 2003, 2006, 2015, 2018, 2019 (Generally, 1930, 1939, 1940, 1941, 1945, 1958, 1963, 1997, 2002, 2004, 2009, 2014 were not bad for heat waves) - Currently evaluating these years for winter: 1951-52, -1957-58, 1963-64, 1965-66, 1972-73, 1979-80, -1980-81, 1982-83, 1986-87, 1991-92, 1992-93, -1993-94, 1994-95, -1995-96, 1997-98, 2002-03, 2003-04, 2006-07, 2012-13, -2014-15, 2015-16, 2016-17, -2017-18, 2018-19, 2019-20.

It's interesting looking at the warm ENSO events here with very hot July weather following cool-ish June weather. Not real common. I reckon we finish the month at 94-98 degrees for the high, +2 to +6. July has very little variability here long-term, so persistence without rain is necessary for that level of heat. Year Jun/Jul/Aug 2003 88.4 / 96.7 / 90.3 1951 89.6 / 96.2 / 88.8 1979 89.1 / 96.1 / 90.9 1963 90.3 / 95.2 / 87.9 The four great warm ENSO heat waves tend to collapse here to near average in August. We'll see how that goes. Looks familiar? Those four years combined would be a pretty interesting winter blend. Just about everyone would be happy - heat, warmth, it's all there in different places at different times. But I don't buy the blend. I do think it's increasingly likely that both 2003-04, and 1951-52 are among the better winter analogs. Will probably combine those two with one or two strong El Ninos, and subtract out cold ENSO events with a +PDO to fix the PDO.

Here is an example of the anti-ENSO v PDO looks I mentioned above. July 1-11 1983 (La Nina following El Nino, rapidly declining solar, +PDO, after a traditional net-aerosol gain volcanic eruption, i.e. net cooling.) About as opposite as possible. It's a cold/warm/cold split left to right in early July 1983. 1951 if you assume +1 to +2 of warming given the ocean gain in heat content, is warm/cold/warm from left to right, like this year.

I've been experimenting with the incoherent PDO / ENSO set of winters. In other words: (El Nino w/ -PDO) - (La Nina w/ +PDO). Just to see what it looks like. Still not many good matches. If you're a bit generous with "anti" PDO years. You get something like: 1951-52, 1953-54, 1963-64, 1965-66, 1972-73, 1994-95, 2004-05, 2006-07, 2018-19, 2019-20 for -PDO / El Nino. and 1983-84, 1984-85, 1985-86, 1995-96, 1996-97, 2000-01, 2005-06, 2016-17 for +PDO / La Nina. The strongest -PDO v. +ENSO years are 1951-52, 1953-54, 2019-20, and 1983-84, 1984-85, 1995-96 for -ENSO. So... 1951-52 (x3), 1953-54 (x2), 1963-64, 1965-66 (x2), 1972-73, 1994-95, 2004-05, 2006-07, 2018-19, 2019-20 (x3) minus 1983-84 (x3), 1984-85 (x2), 1995-96 (x3), 1996-97, 2000-01, 2005-06, 2016-17 (x2). This isn't a forecast, it's just to see how the unusual opposites blend, since PDO/ENSO opposite winters occur 1/3 of years or less. The blend looks like the 1972-73, 1997-98, -2014-15 blend I used for June doesn't it?

The best part of not living in the South is when we have strong high pressure/heat in the Summer...we still have mountains. Big mountains. This was 7/9 when it was near 100 in ABQ. Angel Fire is 8,600 feet up - I've not been terribly impressed with the high temperatures with the current ~600 decameter non-sense we get semi-often. But the lows have been a lot warmer than normal - mid and upper 70s. That actually is pretty rare here especially with dew points still mostly in the 30-50 range. We'd typically fall into the 60s even on a 100 degree day.

PDO held firmly negative in June. https://oceanview.pfeg.noaa.gov/erddap/tabledap/cciea_OC_PDO.htmlTable?time,PDO 2023-01-01T00:00:00Z -0.92 2023-02-01T00:00:00Z -1.1 2023-03-01T00:00:00Z -1.63 2023-04-01T00:00:00Z -2.18 2023-05-01T00:00:00Z -1.69 2023-06-01T00:00:00Z -1.9 Since 1950, the developing El Nino years following multiple negative ENSO events are 1951, 1957, 1963, 1968, 1972, 1976, 1982, 1986, 1997, 2002, 2009, 2014, 2018. 1951, 1963, 1972, 1976, 1982, 2002, 2018 were still very negative PDO looks in June (-0.5 or less). Others were within 0.5 of 0 or positive. 1951-52 remained pretty negative for the PDO into Nov-Apr. The other years flipped to near neutral or positive for Nov-Apr. I was playing around with SSTs again today. 1980-81 is pretty solidly opposite of forecast conditions. This is one of the better matches I've come up with for trying to match both current conditions and what the Canadian has in winter. I do expect the warmest anomalies at the surface to be at 120W though, Canadian has 135-145W. Cold pockets are kind of there west of Australia and North America like now. 1979 and 2003 are both notorious for strong Summer heat waves (I believe we had 20 days+ 100F or hotter here in 1979 - at 5,350 feet up - average is three) in borderline El Nino conditions. Blend is decent on ENSO order and solar conditions (pretty high blended).

The link for my analog scoring model seems to have died pretty quickly after I uploaded it before. This one should last for 30 days if anyone wants it. You literally enter what you think will happen on the right and it automatically scores all the years since 1931 as matches on seven factors. The data for Solar, AMO, PDO, ENSO, Modoki, ENSO prior, etc is in there with weightings. Also allows you to quickly see how matching years would change if most conditions were similar, but one of the variables changed. https://easyupload.io/zajyq9 https://t.co/XMMwn7g23u I still like a ~ +2.0C peak or Oct-Nov, steady or falling starting in Nov-Dec, and only around +1.5 for DJF (28.0C) in Nino 3.4. In prior El Ninos following three cold ENSO years, the PDO has flipped mid-year. Typically May-August. So if it is going to happen, it should be anytime now. Still looks negative to me. The fluky flooding in the NE US reminds of Agnes in 1972 by the way. I've been looking for some kind of heavy-rainer up there, whether tropical or not.

Currently, the El Nino of 1951-52 is my top objective match locally to how highs have behaved out of all years since 1931. 1982 is actually not a bad match either among the stronger El Nino set. For July 1951 the high was 96+. I could see that happening here this year - it's been hot this July so far, with all days for the foreseeable future set to be above average highs (92F). The big El Nino of 1982 was volcanic, with an unusually timed big heat wave in Mexico in June (the heat typically peaks there in Apr-May). March 1952 is really cold here, like 7 below average cold. It's sort of similar to last year locally, normal-ish temps winter, with Oct/Nov/Feb/Mar much more severe than Dec/Jan I think most of you would agree Boston has a pretty different climate to Albuquerque (above) but 1951 also shows up as the top match for Boston year to date. These are the El Ninos in the Boston & ABQ top eight matches blended (1951-52, 1951-52, 1969-70, 1982-83, 2002-03, 2006-07, 2019-20. It's a pretty mild overall cold season, but Oct/Nov are stupid cold in most of the US, with March also very cold. That's consistent with my research on transitional seasons in volcanic years. Off the top of my head - this would be what 35 inches of snow for Boston and 10 inches where I am? Actually would bet it's closer to normal snow in DC/Philly than Boston but I'd have to look.

Here is an updated look at how Nino 3.4 is behaving. Nino 3.4 is currently pretty close to 1987 (28.65), 2002 (28.44). Still well below 1997 (28.82) and 2015 (28.90). 1957 should start to fall off as a good match for a while if Nino 3.4 stays in the mid-28s. 1957 26.04 26.54 27.46 28.23 28.55 28.36 28.17 27.69 27.44 27.42 27.62 27.90 1972 25.62 26.30 27.09 27.89 28.32 28.18 28.14 27.95 27.95 28.26 28.61 28.69 1997 26.01 26.38 27.04 27.98 28.58 28.82 28.86 28.75 28.85 29.08 29.12 28.89 Mean 25.89 26.41 27.20 28.03 28.48 28.45 2023 25.83 26.30 27.19 27.96 28.40 28.54 -99.99 -99.99 -99.99 -99.99 -99.99 -99.99

I've uploaded my main analog model here for anyone interested with data through 2022-23. It took me years to develop, but without the AMO updating any more, it's not going to be usable the way I've used it. It's still very powerful as an analog tool though. It basically allows you to do regressions without math, i.e., what if we had the exact same conditions, but the Atlantic was cold? What if we had the same Atlantic/Pacific, but the El Nino was weaker, etc. What if we had the same El Nino and Pacific and Atlantic, but the prior ENSO event in winter was different? I've included how I calculated / sourced the variables to the right of the data as well. If you want to help me out, you can tip me on PayPal - I have quite a few other things I can share or develop if anyone tips. https://t.co/wsurUGcxYv https://file.io/vfeFLkdfbK1e

Subsurface reading will change from a prelim to final number shortly. But for now 100-180W was +1.46 in June. The 100-180W subsurface was warmer than all Junes except 1997 and 2015 back to 1979. https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ocean/index/heat_content_index.txt Apr/May/June: +1.19/+1.11/+1.46 for 2023 This is a decent match overall. Although not by "tendency" since it misses the weakening in May in 2023. Subsurface: Year Apr May Jun 1982 0.96 / 1.01 / 1.11 1997 2.17 / 2.01 / 2.25 1982 0.96 / 1.01 / 1.11 2009 0.65 / 0.87 / 1.13 Mean 1.19 / 1.22 / 1.40 2023 1.19 / 1.11 / 1.46 1982 has behaved somewhat similarly in the subtropics, with a rare El Nino Central Mexican heat wave in June spreading to Texas. Actually think this is an OK blend just for the Nino zones. Looks very different/wrong for the Indian Ocean / North Pacific and North Atlantic though. That 120W center for the warmest waters against local averages keeps showing on the models. 1982-83: 28.79 (+2.29C DJF) 1982-83: 28.79 (+2.29C DJF) 1997-98: 28.87 (+2.37C DJF) 2009-10: 28.14 (+1.64C DJF) Blend: 28.65 (+2.15C DJF) - I'm expecting ~28.3C for winter, +/-0.2C

My issue with 2014-15 and 1957-58 has always been how ideally positive the PDO was in those years. PDO should continue to weaken toward neutral. But I don't really expect it to go super positive. I expect it to stay negative through at least most of the Nov-Apr period. If nothing else with 1957, the timing of the MJO looks off to me. Major systems in the Gulf in June like Audrey usually occur with specific MJO setups. 1957 is relatively opposite June isn't it? Cold middle US, warm East & West. 1986 June is super cold Texas, super hot Southeast and Western US - also pretty opposite. Both consistent with opposite PDO conditions. The weeklies for June imply Nino 3.4 was at 28.6C - in which case we're closer to 1997 and 2015 than anything else, at least for a bit.

The models are finally realizing this event is not going to be historically strong - this has been obvious for a while by matching actual SSTs. The Pacific looks a lot more realistic to me on the new run with a weaker El Nino. The temperature / upper level maps at this point last year were completely wrong for the actual winter pattern. But the SST patterns start to be pretty realistic at this time for winter on the models.

One nice thing about the dry heat early is it makes my burgers taste better later in the season. 1957-58 (286 sunspots/month July-Jun) and 1986-87 (19) are both bad matches for solar individually, but the blend is solid (152). June 2023 is around 150 sunspots. I believe the new Canadian will be out tonight. But 7/1 is Canada Day right? So might be a day late for that.

1957/1986 is pretty good for solar/strength/order but I don't like it for the IOD and AMO setups. You have to get pretty cute to match the main features the way the Canadian has it. This is one of the blends that's kind of a C+ match. But it's hard to get the western Pacific right with the warmest waters in the Nino zone as shown on the Canadian. 10 blend analogs, 3 blend anti-logs 2015-16 x3 2018-19 x3 1972-73 1997-98 1992-93 2012-13 -2014-15 x3 That's a very similar winter to last year if it developed. But I don't think the June Canadian run has quite the right idea, and my blend isn't a great match anyway.

I've been trying to match historical blends to what the CFS and June 1 Canadian run show for winter SSTs. Even with anti-logs in there I've not found a blend I'm satisfied with yet. My guess is the new Canadian run out tomorrow will have a better handle on most features now that we're only ~five months out from the start of winter. Potentially only ~4 months out from the event peak too. I know the seasonal forecasts showed a weak monsoon out here. But Beatriz in the Pacific should dump a lot of moisture into the monsoon circulation when it scrapes Jalisco and the Sea of Cortez. Monsoon should gain power quite quickly after. By the way, the AMO time series that went back to 1856 is no longer being updated. Not sure why exactly. Not thrilled about that - the Atlantic looks like the dead cat bounce years in the early 1960 before the AMO flip. https://www.psl.noaa.gov/data/timeseries/AMO/ I am increasingly optimistic for a snowy winter for a large area of the US. The current guess is a severely cold winter somewhere in the SE US, with a pretty warm winter Northern Plains. I've been toying with doing a weak+strong El Nino blend minus the recent La Ninas with +PDO features, with 1992 and 2012 added in for volcanic legacy and low sea ice. For me to go gung-ho on 1972 as the main analog, I'd like to see the NE US get hit by 1-2 tropical storms or hurricanes ala Agnres and Carrie.

Yeah...no monsoon just yet. Not quite enough moisture. ABQ - D a t e Time (mdt) Wind (mph) Vis. (mi.) Weather Sky Cond. Temperature (ºF) Relative Humidity Wind Chill (°F) Heat Index (°F) Pressure Precipitation (in.) Air Dwpt 6 hour altimeter (in) sea level (mb) 1 hr 3 hr 6 hr Max. Min. 24 17:52 W 18 G 29 10.00 Fair CLR 94 -3 96 81 2% NA 90 30.08 1009.4 24 16:52 W 14 G 25 10.00 Fair CLR 96 -0 2% NA 91 30.09 1009.7 24 15:52 S 8 G 20 10.00 Fair CLR 92 3 3% NA 88 30.12 1010.8 24 14:52 SW 9 G 18 10.00 Fair CLR 91 4 4% NA 87 30.13 1010.9 24 13:52 Vrbl 7 G 22 10.00 Fair CLR 88 5 4% NA 85 30.16 1011.6 24 12:52 SW 10 G 17 10.00 Fair CLR 86 7 5% NA 83 30.17 1012.0 24 11:52 SW 5 10.00 Fair CLR 81 11 83 57 7% NA 79 30.19 1013.6

No real sign of the monsoon yet for us. The heat wave in the Central Mexican highlands is starting to break though. That often precedes wetter periods here by 1-3 weeks. Here is one of the towns in Hidalgo that averages 75/51 for high/low temps in June at 8,000 feet. "Cool" to see 40s showing up in their forecasts again. I know people worry about the "global south" with climates warming, but mountains are still mountains. We should do our annual post Summer solstice flirtation with 100 degree temps here. Fortunately, the dryness remains, so lows are still between 58 and 70 most of the time. It's only in the 90s even on a record hot/ near record day here for 6-12 hours, even right at the solstice.

CFS is actually showing a fairly similar pattern that looks similar to the July 1972, 1997, -2014 blend that has worked fairly well for June. Some coolness NW US, pretty hot elsewhere. Mexico is quite sensitive to Nino 3 / 1.2 warming. From what I can see, the current heat wave over Mexico is somewhat similar to 1957, 1982, 1991, 2009 since 1950. By severity and timing, 1982 is the closest El Nino June heat wave match. I look at some of the towns in Hidalgo that are 7,000-9,000 feet up without major heat island effects. Many of those towns typically run 75F/51F high/low ranges in June, but are set to finish at like 83F/53F for the month (which to be real, is still quite lovely). So the 63F average temperature for June v. 68F normal is consistent with the +5F shown by Hidalgo state (20N/100W). The other common June El Nino pattern is below - Guadalajara, at ~5,000 feet is really roasting (this is v. average June highs of 87F / with a 63F low). As with Hidalgo though, the absence of rain / humidity / clouds enhances the daily highs but keeps the lows cold or near average. The records in a lot of the towns in Hidalgo are in the low 30s for lows in June-August. That's because you still have 11+ hours of darkness and high elevation. Add dry air to that and the lows can really plummet once in a great while. Guadalajara is thus seeing cool lows and very hot highs at the moment for June, for +4F or so net..

It's interesting to look at the stronger El Ninos by solar conditions. We've not had a major El Nino (27.5C+ in DJF) with high solar (>55 sunspots on average for July-June) in a very long time. 1940, 1957, 1972, 1982, 1991, 2002. There are a bunch of bubble years - 1997-98 and 2015-16 for solar, and several others for strength. The major El Ninos with low solar are 1965, 1986, 1994, 2009 Not a big deal for spatial temperature patterns. Both are warm north cool south looks. Pretty different precipitation though. You can see a smile shaped subtropical jet pattern in the high solar years. Different storm pattern to a SW->NE running subtropical jet that hits Baja California to LA but misses most of California and the Northeast in low solar years.

Here is one of the blends I've been watching for a little while. Will this happen? Doubt it - but we'll see how July-November goes. The blend probably needs to be warmed up 1-2 degrees nationally. Should be cold in the West in July if the 1997, 1972, -2014 blend were to hold. October would be severe again - and very cold nationally. That's consistent with my volcano ideas. Pretty cold December nationally too. East then torches Jan-Mar.