Analoging in the 2020s

[raindancewx]

This decade should feature some pretty extreme weather from both natural climate cycles and the Earth warming. I have a couple thoughts for the decade overall:

1) It's increasingly rare to have a season that isn't warm or wet. At a local level, I think the wet signal is as likely to be dominant as the warm signal in any given season. You don't always have a warm/wet season...but it's pretty rare to not have at least one now. Looking back at my last 30 winters (1989-90 to 2018-19) in Albuquerque, using the 1951-2010 averages for highs (which are much less impacted by Earth warming than the rapidly rising lows), and precipitation, I think only four winters weren't warm or wet or both - 1989-90 (cold/dry), 1996-97 (cold/dry), 2009-10 (cold/dry), 2012-13 (cold/dry). I think this is a good generalization to the US overall.

2) Data mining is useful for re-creating tropical inputs not available in the old days (MJO, OLR, etc). I find that re-creating near exact matches to precipitation patterns for an extended period is quite predictive for several months in a specific location, even if the analogs used for the composite are 50-100 years old. Precipitation patterns likely tell you about the position of highs/lows and MJO input, and how it is all interacting.

3) The AMO is fairly likely to go negative this decade, but it will be a 'warmed up' cold phase. I think the flip could be as early as this year, but more likely mid-decade. I'd expect to see more snow deeper into the SW, SE, and Mexico, like in the late 1950s and 1960s as this transitions happens.

4) Old analogs are likely to still work in El Nino years and Neutral years if the AMO/PDO match too, but are not likely to work in La Nina years. A lot of La Nina years in the old days had extremely cold Nino 3.4 and Nino 4 values, which have not been witnessed in recent La Ninas, and were surrounded by cold in the subtropics. The strong El Ninos in the 1930s-1950s act a lot like weak El Ninos today, due to similar SSTs since CPC uses a sliding scale that changes every five years for ONI. A strong El Nino in 1930 had SSTs around 27.5-28.0C, not much warmer than 2018-19. The strong La Ninas used to have SSTs below 25.5C and low SSTs around them - even the strongest La Ninas now don't typically get surrounded by cold SSTs. The strongest El Ninos did have fairly warm SSTs surrounding them in positive PDO years. 

5) The solar cycle is a pretty big wild card. Long-term, low solar promotes cold in the interior West on an annualized basis, since we have the most sunlight and dry air, and are furthest away from ocean moderation even though it is more of a cold signal in the East in mid-Jan to mid-Apr. Low solar has some impact on precipitation timing and development too, at least if you look at it statistically. Certain "high solar" events are not likely to appear much, or at all, this decade. If we're going to get a tendency where 40% of winters are El Nino/near El Nino (27.0C+), and 70% of years are low solar, I think we'll see some pretty shitty winters in the NE corridor for snow totals for a little while. The low solar + El Nino thing tends to not work out for the immediate coast for snow. It can be a great pattern during La Nina though.

6) I'm expecting a major volcanic eruption (VEI of 4 or higher) in the tropics by the end of the decade. If this coincides with the AMO flip, you will see a fairly prolonged return to slightly better sea ice extent in the Arctic (1997-2006 levels) during September, and you'll see some increased frequency in -NAO months during Nov-Apr compared to the last 10-20 years. 

I'm sure some of you will disagree with this. I could be totally wrong, but that's my sense of the background trends for the decade.