Don't Rely On A Rodent's Weather Forecast
SPEAKERS
Jim du Bois, Kenny Blumenfeld
Jim du Bois 00:00
Predictions, prognostications, and Punxsutawney Phil? This is Way Over Our Heads. It's a weather and climate podcast. I'm Jim du Bois, Kenny Blumenfeld's a climatologist. Kenny, how are you doing on this Groundhog Day?
Kenny Blumenfeld 00:15
You know, it's a nice brisk, cold morning. Rode into work on the bicycle and didn't regret it. But understand why people would if they didn't have the right gear. How are you, Jim?
Jim du Bois 00:26
You know, Kenny, I'm doing well. I agree with you. It was a little chilly this morning. But I found it kind of bracing in a way when I stepped outside. So, it gave me energy. I don't know if...it sounds like from what I'm telling you, Kenny, maybe, maybe I'm starting to have a little affinity for winter.
Kenny Blumenfeld 00:44
Well, I mean, you know, it's about time. You've been here how many years?
Jim du Bois 00:48
My entire life. But...
Kenny Blumenfeld 00:49
Yeah. Well, you know it's good. At some point, you stop fighting and you just accept.
Jim du Bois 00:55
Right, right.
Kenny Blumenfeld 00:56
You know, you gotta live with this sucker for five months out of the year, most years. So, you know, you can either enjoy those five months and make peace with the fact that you know, you might need some extra layers. Or you can...I can't imagine the alternative. Can't imagine spending five months in...
Jim du Bois 01:18
In lockdown?
Jim du Bois 01:20
Yeah, misery.
Kenny Blumenfeld 01:21
I can definitely imagine five months in lockdown. Maybe previously I couldn't imagine that.
Jim du Bois 01:26
Exactly, exactly. Well, Kenny, I'm doing my best to embrace winter and I really need to do that because if you are to believe Punxsutawney Phil, he is the groundhog who comes out from his den as legend has it every February 2. He's done so since 1887 in Gobblers Knob, which I just love the name of that. That's just fabulous. But Punxsutawney Phil emerged today, saw his shadow, which means if you believe in such things six more weeks of winter. So that's why I went out this morning and I opened my arms and just felt winter all about my body because I know if Punxsutawney Phil is correct, I better get used to it for another six weeks.
Kenny Blumenfeld 02:12
I mean, I was out this morning, and the Sun was starting to shine, but I couldn't find my shadow. I wonder what that's about?
Jim du Bois 02:19
Well, apparently, from what I understand, it doesn't always matter if the Sun is out or if it's an overcast day because I guess on occasion, Punxsutawney Phil has seen his shadow on days where according to the climate records, it was overcast in Punxsutawney, Pennsylvania, and there have been other days when he has emerged from his den and not seen his shadow but according to the climate records, it was a bright sunny day, so I guess it all depends. But anyway, we almost didn't have the traditional Groundhog Day this year because Milltown Mel who apparently is the groundhog who has served as Punxsutawney Phil for multiple years, passed away recently and as of yesterday, according to NPR, they were scrambling to find his replacement and weren't sure that they would in time for the usual event to take place this morning. But they found a stand-in. Punxsutawney Phil saw his shadow and the rest is history.
Kenny Blumenfeld 03:22
I'm gonna, I'm gonna just do a timeout here.
Jim du Bois 03:24
Okay. Yeah.
Kenny Blumenfeld 03:25
Real quick, Jim. I think that Milltown Mel is a different, that's a different groundhog.
Jim du Bois 03:25
Oh, it is...I totally, okay maybe I confused myself.
Kenny Blumenfeld 03:32
I think it is. I think, I'm on the NPR story. And it's basically there's several different cities that have kind of their version of Punxsutawney Phil.
Jim du Bois 03:43
Oh, you are correct.
Kenny Blumenfeld 03:46
Punxsutawney Phil is alive. It's that Milltown Mel croaked.
Jim du Bois 03:51
Well, then that's my mistake because now as I skim further in the story, I see that he is associated with Milltown, New Jersey, and has absolutely nothing to do with Punxsutawney, Pennsylvania. So, I digress. But I guess it serves the point that apparently this phenom of the ability of groundhogs to predict the winter weather isn't limited to just one city in one state. This is, this is a, perhaps it's even a worldwide phenom. We don't know.
Kenny Blumenfeld 04:23
Right. And then the Groundhog Day tradition, I think that Punxsutawney Phil is the longest-running one in the US. You know, where this particular groundhog became the winter spokesperson for the rest of the country is beyond me. But let's just be honest here, Jim. With the forecast being either a mild winter or six weeks more of winter, on February 2, to Minnesotans, that's a little like saying okay, you can have vanilla ice cream, or you can have the one that's not strawberry. If we only have six weeks of winter, that's actually a pretty warm winter. So, to us, what's the difference?
Jim du Bois 05:05
Good point, Kenny.
Kenny Blumenfeld 05:07
What's the difference? If our winter's over on March 15, we got off easy.
Jim du Bois 05:12
Exactly, exactly.
Kenny Blumenfeld 05:13
And that's even in light of, you know, the way that the climate has changed. Certainly, we have had winters that have ended before March 15. But you know, you also got to remember the Thunder Blizzard 2018. We had heavy snow in mid-April in 2019 and 2020 in Minnesota, and some listeners may recall that on May 1, May 2, somewhere in there, in 2013, we had over a foot of snow in southeastern Minnesota in the area, you know, around Northfield and Rochester in towards the Red Wing area. So yeah, if we only get six weeks of winter, good for us. And I don't want to pick on Phil, I mean, this brings up a lot, a lot of questions. Jim. Right? A lot of questions. Okay. So, if a groundhog, does a groundhog actually see its shadow? And if a groundhog sees its shadow does it know it sees its shadow? Do groundhogs know what shadows are?
Jim du Bois 05:59
Yeah, you know, I think the problem is you just...
Kenny Blumenfeld 06:25
I think it's a fair question.
Jim du Bois 06:26
It's a very fair question. And I guess the followers of Punxsutawney Phil just go on blind faith because I believe his handlers are known as the Inner Circle. It's a very elite group of people in Punxsutawney, Pennsylvania. They're the guys who wear the top hats that you see with Punxsutawney Phil. And one of them is able, apparently, to speak Groundhogese. So, it's not only that Punxsutawney Phil does or doesn't allegedly see his shadow, he also engages in a conversation with this guy from the Inner Circle, and apparently goes into greater detail about his predictions for the weather. So, who knows?
Kenny Blumenfeld 07:09
You know, I figure while he's at it, they should get his opinions on other things.
Jim du Bois 07:13
There you go. Maybe he can fix the political mess we're in?
Kenny Blumenfeld 07:17
Yeah. Or, you know, who's gonna get the most golds at the Olympics? I, I think that you know, I'm glad that they do this because it is a, it's a cherished tradition. Listeners should know it is total malarkey. The National Weather Service did an assessment of recent forecasts on a national basis. So, this was basically during the 2010s. Punxsutawney Phil's batting average was just under, I think it was coming in around 400.
Jim du Bois 07:56
Not even a coin flip.
Kenny Blumenfeld 07:58
Yeah, right. So basically, this notion that the groundhog sees its shadow and makes an accurate prediction about the remainder of winter is actually less accurate than random chance. So, okay, so have fun, enjoy it. I would not make plans specifically based on it. But you know, in a way this gets at kind of our larger need, this larger desire that we have to understand things that are out of our control. And the weather, especially late in winter, and early in spring is one of those things. You know, people can feel better about the planting season coming up if they understand that the weather is going to be favorable for it. And so, it's a nice tradition. I just hope people understand it's completely non-scientific. And I think most people do get that.
Jim du Bois 08:54
Absolutely. And we should point out, too, that the science of predicting the weather is, I guess, in the scheme of human existence on this earth, a relatively recent phenomenon because, correct me if I'm wrong, Kenny, but we really didn't even understand what warm fronts and cold fronts were until the early part of the 20th century, correct?
Kenny Blumenfeld 09:14
Yeah, yeah, that is correct. And, you know, in Phil's defense, there are professionals with advanced degrees in atmospheric science who try their hand at seasonal forecasting, and I can't say that they're that much better than Phil. I mean, you know, they, obviously there's nuance, and they, this is the Climate Prediction Center, but they're not the only ones, they're the most famous ones through the you know, mainstream will probably recognize those maps that say it's going to be above normal or below normal for temperature, precipitation. There are also private sector, long range forecasters who work for, you know, they work for the various industries that would really care, energy sector and agricultural sector. There are commodities forecasters who try to, you know, get a good sense of what's going to happen in the next several months or in the coming season. It's hard work. I mean, that's just, it's really hard. And it's not only because forecasting the weather in general is hard, it's just because it's, it's such a, it's such a complex system. I mean, and think about the things that we're looking for. I mean, so let's say, Jim, we're talking about the weather three months out, and there's a strong indication that it's going to be a cool period and a wet period. And this is hypothetical. But now, just imagine that, that it's largely correct. And it is, you know, generally, below normal temperatures, and fairly wet, but it's punctuated by an astronomically hot, you know, 15-day run of, you know, warm and very dry weather. Okay, so then, was it right? Was the forecast, right? You know, was it? There's so much nuance in here, too, and what people experience and remember, is often different from what the facts are. So, for example, last winter, the winter of 2020, to 2021, making sure I get my years straight here, it was a warm winter. It was, you know, and we were actually on our way to one of the warmest winters on record, probably top 10 territory. But there was a two-week interlude during February where it was very cold. And it, you know, you and I talked about this, but temperatures were 20 degrees, 25 degrees below normal on average for a period of almost two weeks. And so that's actually what people tend to remember about last winter, and they don't so much remember the fact that it was actually, you know, 80% of the season was mild. They don't remember that so much. They remember the piercingly cold outbreak in February. So, it's interesting. It's hard work to make those forecasts, and you're right, just forecasting in general, is really a, I mean, you know, the old joke is that Aristotle coined the term meteorology and, you know, within a couple hours, someone asked him what the weekend was going to be like. But truthfully, forecasting is a relatively recent endeavor, and really got kind of serious in the 20th century.
Jim du Bois 12:45
And we should point out that the accuracy of forecasts has improved dramatically, particularly since the 1950s. We actually have some stats. I have one here that in the mid-50s, the 36-hour National Weather Service--then the US Weather Bureau--forecast was about 20% accurate. By 2014, that 36-hour forecast was accurate over 80% of the time. Very similar for the 72-hour forecast. In 1975, 25% accuracy. 2014, 70%. So, we've made major strides in the last 40 to 60 years. Is this, Kenny, due to the fact that we now have supercomputers that can crunch numbers, and we didn't have access to that prior to say the mid-1950s.
Kenny Blumenfeld 13:37
Yeah, that's a great point, Jim. And I would say that it's a combination of things. I mean, one is forecasters are now better trained. Meteorology is still pretty much the same science, but the tools that meteorologists have to bring into forecasting, they're not just super computer based, but there's just other analytical tools, and there's concepts that really weren't available in the early decades of forecasting that are. So, you have some very well-trained forecasters who have some very good tools. We have much better, in many cases anyway, better monitoring and better data, better data quality, better programs to help analyze the data. So, it's just all in all a better system. And, you know, some of the stats that you read, I think it would be easy for a person to say, well, yeah, but they're still wrong, you know, X percent of the time. But there's been, you know, in our lifetime, and certainly in the lifetime of people who are older, there's been a huge change in the kinds of mistakes that are sort of acceptable in forecasting. And one thing that you know, some of the younger listeners might not even be aware of is that it used to be really common for a forecast on a day to, you know, completely miss the potential for some kind of a cataclysm or a disaster and you know, it wasn't like it happened all the time. But sometimes, you know, forecasters would know that it could rain today, but they didn't foresee tornadoes, or they said that it would be partly cloudy, but you ended up with a flash flood. And even in the 1980s, we had some phenomena where it was not supposed to snow, and then we got a lot. Or in 1984, there was this surprise blizzard that really wasn't particularly well forecast in western Minnesota and killed a lot of people. And those kinds of things really don't happen anymore. And now, there's more of the other kind of error where maybe people are expecting severe weather or expecting floods, and they don't materialize where the snowstorm, you know, is forecast to hit, but instead, it ends up 40 or 50 miles away. And that can really upset people because they make plans based on that. But it's a totally different kind of error from what we used to get. I mean, now, you know, forecasters largely can see most of the big weather events several days out. And they generally make accurate forecasts on them, with the exception of, you know that very close geographic specificity that's still kind of hard to resolve. And so ,you still see those types of mistakes. But we're sort of done with the days where, you know, we expected it to be clear, and we ended up with huge downpours and damage, or the days where, you know, the forecast called for, you know, brisk winds and temperatures in the 20s but people ended up freezing to death with temperatures plunging, you know, well, well below zero. So, big changes, and you know now, the forecasters are often, they often have a pretty good handle on the general picture five days out. And when I was a kid in the 80s and 90s, early 90s, that was, you know, five days out was like a holy grail. It was a best guess. And it was definitely a guess at that time. And so, there's been huge improvements in forecasting, indeed.
Jim du Bois 17:09
Kenny, is it fair to say that forecasting may be a bit more challenging in areas with high variability, and I'm thinking like the Midwest, and maybe easier in places where there are large moderating influences like areas around an ocean. Any truth to that?
Kenny Blumenfeld 17:27
There's some truth to it. But I would be careful. There was a huge snowstorm this past, I guess it was the past weekend in Boston.
Jim du Bois 17:35
Yes, two feet of snow.
Kenny Blumenfeld 17:37
Two feet of snow. And that forecast was definitely a razor's edge type of forecast. You know, there's this big, explosive low pressure system, tracking over the ocean, not too far away from Cape Cod, you know, 100 miles or so from Cape Cod. But its position really, really mattered. Because that determined where the precipitation cut-off would be, and where the precipitation, you know, wouldn't go. And so with those sort of razor's edge forecasts, we've seen, you know, areas in the East Coast that are expecting a foot, foot and a half a snow, and they end up with six inches, and everybody's mad, or they end up with two inches. So, they do have some of the same issues as us. But the ocean does produce a moderating influence. I would say one of the bigger factors that makes the weather easy or hard to forecast is, aside from the variability, which is one piece of it, is the kind of complexity of the general weather patterns. So, for example, Hawaii, doesn't really have very much complexity. Most days, the Sun comes out, especially on the sunny side of the islands, they've kind of a sunnier side and the cloudier side, depending on if you're up the slope, going up, the slope with the winds are going down the slope with the winds. The winds in Hawaii are usually out of the east. So usually, the west sides are very, very bright and sunny and dry, and the east sides are kind of wet. In any case, the weather there doesn't change all that much. And it's fairly simple. They don't get low pressure systems frequently, they usually just have kind of random heating from, you know, the topography and from the patchy sunlight that's there every day. So, you can pretty accurately forecast in most of Hawaii almost any day of the year, you know, sunny until early afternoon with a chance of isolated showers late in the day or by mid-afternoon. And that's a fine forecast almost any day of the year for most of Hawaii. And so, it's a little harder to have huge errors there. And another type of influence that I think is important is, you know, here in the Midwest, and also on the East Coast, we have a lot of these very sharp fronts that move through bringing in, you know, really cold air from northern Canada. And those are those sorts of razor's edge type of situations where there's a sharp cut-off on where the precipitation is and isn't in the path of the weather system is going to be really important to the success of any forecast. But out in the western US, a cold front is a different animal, because when the winds come out of the, you know, when the winds come out of the west in Seattle, they're just coming off the Pacific Ocean. It's not the same kind of airmass transition that we have here. And they tend to get these very broad areas of precipitation that might be, you know, 1000 miles long or wide. And it's just much harder to make big mistakes when you're dealing with a huge mass of ocean moisture. So, I'd say complexity or simplicity of the weather pattern, which does get at the variability. And then whether the kinds of things that you're forecasting are generally very geographically focused. So being right or wrong is kind of a matter of miles. So that's one thing versus these larger kind of broad systems where, you know, you can easily forecast that it's going to rain because there's a swath of rain that's the size of three states moving on shore. So yeah, there's some differences here. And I would say cutting your teeth as a forecaster in the Midwest is probably a good way to get used to the real difficult type of weather where there's a lot of geographic specificity, there's a lot of sharp changes, and as you pointed out, there's a lot of variability. Changes from, you know, week to week, day to day even.
Jim du Bois 21:53
Well, we will talk more about forecasting in a future episode. But for now, I think it's safe to say rely on the science, rely on the professional forecasters, and don't put a lot of faith in a rodent.
Kenny Blumenfeld 22:07
That's a great, that's a great statement. Jim, I support that.
Jim du Bois 22:11
All right, Kenny, great talking to you as always, and we'll check in with you soon.
Kenny Blumenfeld 22:16
Very good. Thanks, Jim.
Jim du Bois 22:17
This is Way Over Our Heads. It's a weather and climate podcast. I'm Jim du Bois, Kenny Blumenfeld's a climatologist. We'll catch you next time.