Clear Skies

A couple of days ago when I saw the forecast for the clear and cold weather that is now upon us, I reflected on the fact that we hadn’t yet had a cold snap this winter. It seems like in most years we get some cold weather in November or December (though I think most of the coldest weather tends to be in January or February). I got curious and started going back through the Weather Underground historical records for Sitka and saw that the last day we had which rated as “Clear” or “Mostly Sunny” was the first of November. Two months without a clear day seemed like a bunch, so I started going back further.

I was a little surprised to find that I had to go all the way back to May to find 2 clear days in a row, and the last time we had three was February. Including mostly sunny days was a little better, but the last time we had 3 or more Clear/Mostly Sunny days in a row was still last May. Of course I’ve never really looked at how many days in a row of Clear/Mostly Sunny skies are typical in any given month, so I didn’t really know if this was typical or not, so I looked back a bit further.

In the end I included 10 years of data 2002-2011. For each month I checked the number of clear days, the number of mostly sunny days, the longest run of Clear days, and the longest run of Clear/Mostly Sunny days. There were some problems with this method, as it’s not entirely clear how Weather Underground decides how to rate a given day. Looking at the daily summaries I saw for each observation (typically at least one an hour, each day), there was an entry for “conditions” which included Clear, Mostly Sunny, Mostly Cloudy, Overcast, etc. It would be better to get that data and do the summarizing myself, according to what I’m interested in, but that will be a much more involved project (which I may get to at some point).

I found that 2011 was one of the least sunny in the last decade. There were only 19 clear days in 2011 (7 of which were in January, with an addition 4 each in February and March). 2008 was second with 20 and 2007 had 21. With mostly sunny days included, 2011 had 56 total – second to 2007’s 49. 2008 had 59. When January is dropped, 2011 falls below 2007 45 to 47.

Of course as often happens, I wandered through the data looking at various other comparisons and summaries. I’m planning to write a follow up post with some supporting plots in the coming days.

It’s interesting to me that I didn’t perceive 2011 to be a particularly gloomy year. I’m not entirely sure why that is the case. I imagine part of it is that I’m used to cloudy weather, so I don’t make particular note of it, even if there’s more than usual. (On the other hand, months/years with more than typical sun do tend to stand out more.) It could also be due in part to how Weather Underground was summarizing the condition reports into a single condition for the entire. For example, a day where a little rain fell at 2am, but it was clear by 8am and remained clear throughout the day would probably show up as a day with rain – even though my experience of it would have been as a sunny day. It will be interesting to see how the general results I’ve found here compare to a more careful analysis with the daily summary data (if/when I ever get that done).

Leave a Reply