Welcome to American Weather


  • Content count

  • Joined

  • Last visited

About raindancewx

Contact Methods

  • Website URL

Profile Information

  • Gender
    Not Telling
  • Location:

Recent Profile Visitors

1,275 profile views
  1. Pretty happy with my winter blend, the same years for winter are almost dead on for right now in the Northern Hemisphere, except in the Indian Ocean.
  2. You are into visual checks right? This is what my analog system (for winter) has for October v. current conditions. I couldn't really get the pattern for winter using just La Nina, you need some cold-Neutrals in there too. You have to imagine an extra 0.1C of warmth everywhere given the analogs are centered on 1979, but are mostly warm AMO years.
  3. I dunno if the "snow" anomalies I'm linking include sleet/freezing rain, etc, but this forum region did pretty well in terms of snow anomalies in a lot of the winters I like. If SSTs mean anything, its a good sign, since the blend is very close to current conditions. Sanity check for analogs in Oct of analog year v. Oct 22 2017:
  4. 27SEP2017 19.5-1.0 24.4-0.5 26.5-0.2 28.4-0.2 04OCT2017 19.3-1.4 24.7-0.2 26.7 0.0 28.7 0.1 11OCT2017 19.5-1.3 24.4-0.5 26.2-0.5 28.5-0.1 18OCT2017 19.5-1.4 23.9-1.1 25.9-0.8 28.3-0.4 http://www.cpc.ncep.noaa.gov/data/indices/wksst8110.for (link) The traditional SOI value is approaching +8 for the last 90 days, so in a pressure sense we're pretty close to Nina status. Not sure if ONI is really Nina like yet, with the warm up early in October. In retrospect, I've realized that none of the strong El Ninos had any kind of El Nino within two Falls of their formation since at least 1930. I sort of knew that in the Spring but the models are usually OK by July when they had an El Nino, but there was a secondary route in the history where a brief warm up in Summer toward El Nino from La Nina goes cold Neutral by Fall. It happened in 1932, which is one of the reasons I like that year. Big El Ninos --> Next El Nino 1930-->1939-40, 1940-41, 1941-42 1940-41 & 1941-42 (double El Nino) --> 1945-46 1957-58 & 1958-59 (double El Nino) --> 1963-64 1965-66 --> 1968-69 1972-73 ---> 1976-77 & 1977-78 1982-83 --> 1986-87 1991-92 --> 1994-95 1997-98 --> 2002-03 2009-10 --> 2014-15 2014-15, 2015-16 --> at least 2018
  5. I was worried for a while that we'd torch here in Oct, when I had it near average here, but it looks pretty good now. I just checked my analogs for winter and the ocean analogs for winter are actually nearly identical to current conditions. Probably one of the best matches I've ever produced for the Northern Hemisphere actually -
  6. I use the Modoki / PDO / AMO values that match what the oceans are forecast to look like, in similar ONI/Solar years, that had similar weather conditions in my area over a long time. That eliminates a huge number of years. I also tend to throw out years on the opposite ends of the spectrum on a given important factor, say the AMO. It takes a super long time to evaluate the accuracy of this stuff on my end because I have to manually look at the misses by state/region and look up the state sizes to come up with a national accuracy figure, but when the seven things I use are guessed/blended together correctly it does seem to work at 55-85% nationally on a seasonal basis for winter. I'm not Nostradamus, so I'm sure I'll be off on one of the figures - the PDO is actually the hardest to predict in my experience since everything resets in October, My system uses a three tiered weighting: AMO, ONI, Solar: x3 each Modoki, ONIp: x2 each PDO, Monsoon; x1 each Matches are then automatically highlighted if they are close enough to get a score of 8 or more out of 15 possible points. Some years are never fully reproducible, I couldn't get 2014-15 with my system because it had the highest PDO value since 1900 for Nov-Apr, and I basically didn't know what to do. The system is basically designed to rank the years from best to worst match, and then it is much easier to analog everything. It takes out 80% of the time/effort. My analogs are in the winter outlook I put out - https://t.co/ZnvyQletct 1932 (x2), 1943 (x3), 1944 (x3), 1996 (x3), 2005 (x3), 2007 (x3), 2008 (x2), 2012 (x1)
  7. My analog system was able to produce a close match (~80%) to your marked coordinates, hopefully will stay fairly similar in winter. Some of the years are ancient, so you have to imagine an extra 0.2C of warmth in all areas. But this is what I used in my winter outlook that I linked above. It takes into account solar conditions too.
  8. Every time this La Nina seems like it wants to be moderate or even strong (like now) some of the warm water from the subtropics seems to drain into the 5S to 5N area and slowly spill west. The big drop to -1.0C now seems like it will be offset in a week or two by the reds (warm waters) building West again.
  9. I like 1996 & 2007. I don't use the QBO though - but I arrived at those years because the QBO seems to be linked to the solar cycle, and 1996/2007/2017 are at similar places in the solar cycle since it is nominally an 11 year cycle but the current cycle was super weak. 1985, which I also considered, would have been similar too. This month is down to around 10 sunspots, July 1996 to June 1997 had 14 or something? That's how I blend in the sun. I may be mis-remembering this but I've seen a paper with QBO re-constructions from data in Indonesia back to 1900 that implies the weirdness w/ the QBO last year not behaving as expected may have happened in the 1940s too after the triple El Nino of 1939-1942, so 1943-44, 1944-45 (which I have in the analogs) are similar. The QBO is a 28 month cycle right? That means in most solar cycles, it cycles through ~five times. That is one of the ways solar data is useful in my opinion. I don't think this is a cold winter in the Southwest, at least where I am, because forgetting my analog methods, models, etc, the history in the region says hot Junes (which we had) are almost never (15/16) followed by a winter 2F or more below normal - that's true in La Nina, Neutral, El Nino, +AO, -AO, +NAO, -NAO, low solar, high solar etc. So I respect that figure a lot. But enough went right in the Summer and enough has changed in the background state of everything to make me think this is one of our "colder" La Ninas. Still around average though. The average should be fun though...suspect its a lot of slightly warm days and then 20-30 days of incredible cold.
  10. For what its worth.... I suspect the "look" of this October would translate south at some point in the winter with more Arctic air. Maybe look a bit like this?
  11. Will be interesting to see if Nino 3.4 has caught up to the La Nina last year by next week. Still behind as of 10/18. The warm y/y trend has been slowly fading if you animate it though. Suspect the catch up date is around 10/25.
  12. Once the MJO settles down a bit the models probably will be a bit more consistent. It is interesting looking back on the BOM site and seeing the years with similar big amplitude MJO progressions in the Fall.
  13. I don't have much faith in the CFS this early, but it continues to insist on a dry NW for the winter...AND a La Nina.
  14. Also, for what it's worth, this is the latest CFS run. We're close enough to winter that it probably isn't completely out to lunch for the season. I'm not a huge fan of the models for temps/precip, but I do try to look at them. I think they are much better at forecasting ocean temps v. anything else.
  15. That's kind of what happened last year though (decent match to 1942) - pretty dry for a La Nina in Washington, and Montana didn't really see a huge year either. I think the blend has a pretty big Spring in parts of the NW though. To me, there are 27 La Ninas (July-June basis) since 1930: 1933, 1938, 1942, 1949, 1950, 1954, 1955, 1956, 1964, 1970, 1971, 1973, 1974, 1975, 1983, 1984, 1988, 1995, 1998, 1999, 2000, 2005, 2007, 2008, 2010, 2011, 2016. You had substantial dryness in parts or all of the NW in 1938, 1942, 1954, 1956, 1975, 1983, 1984, 1988, 2000, 2008, 2011, 2016 The NW seems to be drier in La Ninas after La Ninas - that's factored in the analog system I have - since I use ENSO order as a variable.