Introducing the May Misery Index

It seems May in Alaska tends to be either really good or really bad, weatherwise. How was your May? Folks in Anchorage had to wait till the last day of the month for the temperature to get above 60 this year. For Kodiak, May offered up only 6 days when the high made or bettered the long term average, and only 6 days without at least a trace of precipitation. Over an inch of rain soaked Glennallen‘s Memorial Day weekend. Around here (Haines) people are still asking each other, in hushed voices, “You think it’s OK to plant yet?” Part of it is, after a long winter, we are hoping, no expecting May to be nice, and when it is, it seems really nice. When it’s not, we tend to take it personally.

So the May Misery Index (MMI) is intended to connect the data to people’s perceptions, stacking up talk such as “good,” “bad,” “miserable” against the statistics, for a reality check of sorts (spoiler: it wasn’t that bad this year). An objective measure to judge Mays, though it carries a subjective (and tongue in cheek) name. It’s really pretty simple, a balance of anomalies of high temperature, number of days of precipitation and total monthly precipitation. Details to follow, but first let me expound on why I chose the high temperature instead of the average.

High temperatures are the most telling at this time of year

My opinion: In summer, the daily high (max) temperatures reveal more about “how the weather is” or “how the summer was,” from a human perspective, than the daily average (average of daily high and low). My first reason is because the when the weather changes on our layman’s scale from “good” (clear, partly cloudy) to “bad” (overcast, rain) the average temperature usually changes little. What happens is the high temperature goes down, but the low temperature stays about the same or often goes up. Here are two examples (click to enlarge):

The blue, sine-like curve is the hourly temperature trace (the dew point and RH have been turned off for clarity). See the daily ups and downs and how this diurnal range varies, shrinking from top and bottom thereby not moving the mean much? Now notice the green bars. This is when it was raining. While clouds alone have the same range-dampening effect, rain usually indicates thicker clouds. The match between rain the daily range on these graphs is pretty strong.

The reason for this is, of course, that cloud cover restricts the dirunal range. Clear weather allows the daytime high to soar due to solar heating, but nighttime lows to dive due to radiational cooling. Clouds block both the solar heating and the terrestrial cooling, resulting in a smaller range: cooler highs and similar or warmer lows than in clear weather. Again, this is for summer: Roughly June, July and August for places north of the Alaska Range; add in part to all of May and September as you move south. The effect is weaker or missing in highly marine locations such as the Bering Sea coast and islands, and the Arctic coast.

The other reason I like to look at high, rather than average, temperatures as a measure of weather “goodness” is that in the summer, the daily low is pretty consistently early in the morning when most of us are still asleep (remember I’m focusing on impacts on humans). Even if you are outside early I doubt you would worry much about how cool the early morning is when the sun is rapidly warming the day into something to rave about. Look at the graph for Tok again as a case in point. When the rain was done and the sky began to clear on the evening of the 29th, the temperature dropped way lower than it had been during the cloudy/rainy few days prior, but look at the highs! The night of the 30th-31st it dropped to 28F. But that was at around 5 am. By 7am it was 41F, and the sun just kept it up all day, achieving a high of 68F sometime between 6 and 8 pm. What are you going to hear when you bump into someone coming out of Fast Eddies after dinner, both of you in shirtsleeves: “Gee it was a lousy day”?

The reality was that the daily average temperature on the 31st, the sunny, 68F day, was a half degree cooler than that of the 27th when it rained a half inch and topped out at 53F.

Side note: Tok’s hourly data shown in the graph is from the relatively new automated station there, a much awaited addition to the hourly-weather network. But not to forget the dedicated coop observer(s) there who have been logging highs/lows/precipitation/snowfall/snow depth once a day, I’ll add that the high-low spread at their station (Tok#2) for May 31st was an even more astounding 44F degrees, from 27 to 71F!

In winter I hold that looking at low temperatures is more telling, but I’ll leave the explanation for another post…

Precipitation has a big impact

The role of temperature in determining how summer weather rates with us humans is actually the smaller part of the picture. Precipitation is more impactful in my opinion. And the number of days with precipitation is more important than the total amount. The lack sun for days on end and the inability to dry out between rainfalls is highly grating on many people’s mental health. So four days with a tenth of an inch of rain each is much worse than one day with 0.4 inches, would you agree? So I weighed this factor heavily in my misery index.

2024 May misery report

Without further ado, here’s a chart of May 2024’s Misery index for a selection of Alaskan cities:

The big picture is that most of Alaska was cool and wet this past month. The way this scale is set up, the higher the number, the more the misery. Negative numbers indicate May bliss. I’ve shaded the Misery index column with light blue for what we might call tolerable misery, 4 or 5, and darker blue for the two on this list that scored 6, pestering misery, let’s say (the scale goes higher, see below). Three or under: no complaining, please; it could be much worse. Below, in the historical graphs, you’ll see it’s been way worse — not sure what adjectives to use for some of those years. So this year, mild to moderate misery across most of Alaska.

This is all new, so I’m just taking a stab. Please chime in with your interpretations, comments, ideas etc.

Details on the index

I kept it simple: no cloudiness, humidity, wind etc, so that it can be used with coop station reports that do not have those data.

My index is relative to the normals of each place. That makes for better comparisons, but I’m not going to say it’s equally useful everywhere. As stated earlier, things don’t work quite the same adjacent to the cold waters of the Bering Sea and Arctic Ocean, so use with caution in those areas.

The index is computed from the three columns inside the bold border, namely the departures from the most recent 30-year averages of high temperature (F), precipitation and number of days with measurable precipitation (>= 0.01″). The first numerical column (ave temperature departure) is the there to further illustrate my earlier point that the average temperature downplays the anomalies of the high temperature, which is what we most experience in spring and summer. In case you are curious here’s the formula: MMI= -0.75*[T(max) departure from normal (deg F)] + 0.5*[precipitation percentage departure from normal] + 0.5*[#days with >= 0.01″ precipitation departure from normal]

I realize there’s a lot of redundancy in the formula: 2 precipitation terms plus the correlation discussed earlier between precipitation and high temperatures. But a close look at some of the data shows the nuances that the MMI brings out. For example, look at Anchorage vs Cordova in the table above. Cooler and wetter (relatively to climate) at Anchorage, but the precipitation came on fewer days than usual, so Cordova’s 9 extra wet days overpowers all that, and by this index it was decidedly miserable in Cordova but a very average May in Anchorage.

How does this May compare to past Mays?

Here’s a time series for each of four key stations in the state, spread between two graphs to avoid congestion:

Lot of ups and downs, but here’s a few observations:

  • 2024 was not too bad when viewed through the lens of history
  • It seems like we’ve been lived through a little less misery as the years have gone by, which is consistent with a warming climate (precipitation changes were weaker influences on the trend.) The entire time series is relative to the most recent 30 year period, which allows us to see this trend. The misery should average to zero over the last 30 years.
  • Bethel and Juneau show larger swings into both misery and bliss over the years and that, I believe, is because they are more maritime vs continental. (Yes, I know Anchorage is on the ocean while Bethel is 50 miles from it, however, Anchorage is mostly surrounded by mountains while Bethel’s terrain is flat and treeless, making it effectively more maritime than Anchorage.) So Bethel is more influenced in May by either cold air blowing off the very cold Bering Sea or warm air from the rapidly warming interior, giving it more variability. Also, when the marine air dominates they can get drizzle, even in the absence of rain-producing storms, which can tick up the number of days of measurable precipitation, a potent part of the misery index. Similar factors apply to Juneau, but to a lesser extent.
  • Those huge spikes on Juneau’s trace to around +12 and -12 for 2012 and 2015 respectively? They are real. In 2012: cold (high temp departure from normal: -8.7F), wet (measurable precipitation on 26 days including 2 with snow!). 2015: very warm (+8.1F) dry (4 days with precip). Another measure: 2012, not a single day broke 60F. In 2015 24 days did and 13 were 70+. Such can be May.

Thanks for reading. I’d love to hear your feedback. Have a blissful June!

Leave a Reply

Your email address will not be published. Required fields are marked *