Why the Weather? If You Know the Past, Why Not? A Case Study
Suppose we take the analogs from last October using the Accumulated Cyclonic Energy index. What do those analogs tell us for April?
They say to watch out for cold across the northern U.S. What did they say for March?
What has March been like?
No doubt it’s been warmer in the South, but we said to look out for a northwest-to-southeast axis of cold in major heating degree day areas, which would be very important to know if you are an energy company, just like snow is very important for retail, city budgets — in fact, everything.
Now what did the analogs say about January, February and March?
Here’s what those months looked like:
What was our January-March forecast issued at the end of September for clients?
The point is that, through researching the past, from many months out we were able to quantify what is being seen now.
Now let’s get back to April.
The idea from six months out was to watch for stubborn cold across the northern states. We know that’s on the table. The blocking pattern that evolved and the knowledge of how long it takes for the reaction to that to work its way out based on past numerous occurrences suggested this would hold into April. The Madden Julian Oscillation (MJO) is behaving in a way that brings closer analogs into the mix, my three favorite being 1975,1982 and 2007.
What did those Aprils look like? Widespread U.S. cold.
Now suppose we go to a blend of the long-standing ideas and these more in-close ideas. What does that give us?
Now watch the climate model’s 10-day runs from 20 days ago, which showed a lackluster April.
There is no cold anywhere. Ten days ago it had this:
Now it’s seeing the cold:
The model is capitulating to the six-month analog blend thanks to a major driver, the MJO. (The two major cold and stormy periods this winter, late December into January and then March, had the MJO rotating through phase 7 then into phases 1, 2 and 3.) Do you see what I’m doing here? I’m using detailed research and analysis of past events — in many cases, well before the CO2 climate control knob craze — matching them up with current conditions and then using modeling to help shore up the forecast!
My argument is that knowing and understanding the past can give you a head start and will mean the models will yield to the analogs if you have the right ones. At which point you can really match things up well.
On the first of March, the climate model showed this for the month:
The combination of the fake spring and forecasts like this probably led to the expectations of the cherry blossoms peaking March 17-20. But this headline just appeared in the Chicago Tribune: “DC cherry blossoms won’t reach peak bloom until around April 10 — 11 days after average bloom.”
The fact is the forecast was a far cry from what happened in March. It’s true that on March 10 the climate model had this:
But it took 10 days into the month for it to show what analogs from six months ago were essentially forecasting!
These are important adjustments if you are concerned about cherry blossoms or if you have to purchase supplies in February or even if you are a hedge fund that’s concerned about a March contract. If you bet on the warm solution, then what? I have been told that, like the winter of 2015, many areas in northern New Jersey are out of salt. Well, it may darn well snow again, and it could be a sizeable storm before this is done, as the 1982 analogs are showing up. Sure, it will melt quickly, but before it does, what are you going to do if you aren’t ready?
Models aren’t worthless. Quite the contrary. They are hugely valuable tools that have something to say, But they are not gods, nor are analogs. But the proper research and understanding of a pattern, be it decadal, yearly, seasonally, weekly or daily, are also huge. For instance, the big nor'easters in early March were shown by us way in advance based on the 1962 analogs. We also honed in on the recent nor'easter based on the 1958 storm.
Think about this: Four days before the March storms the shorter-term models were shoving them to the south or minimizing them. But the analogs said to never surrender and the models eventually succumbed. It takes proper analysis — which takes a lot of ditch-digging work — and model agreement to nail something, But denial of the past leads to folly in the future. My job is to try to show the why before the what, but it takes looking at the past to help explain the future.
Numerical models have no idea what happened before. Good old-fashioned grunt work is essential. Which is fine with me, because, to quote Judge Smails in “Caddyshack”: “The world needs ditch-diggers too.”
I just happen to love being a ditch-digger.
If I can show you what weather is, then you can figure out what weather isn’t. It is essential to know and not deny or change the past — be it in weather, climate or anything else — if you truly want to understand the future.
Joe Bastardi, a pioneer in extreme weather and long-range forecasting, is a contributor to The Patriot Post on environmental issues. He is the author of “The Climate Chronicle: Inconvenient Revelations You Won’t Hear From Al Gore — and Others.”