There were requests from the peanut gallery for more detail in the recent weekly-average-temp graphs. I suspect this was just designed to push me over the edge, but I’ve complied anyway. This took some doing– what with two months of temperature data, taken at 5-minute intervals. But here’s what I’ve been able to produce:
Graph #1: This shows weekly average living room / kitchen temps (blue line), along with associated weekly maximum and minimum temps (green dots). Unfortunately I hadn’t saved the full data for the first week, so no min/max there.
Graph #2: This again shows weekly averages (blue line), but now with orange arrows representing the range of +/- one standard deviation. In other words, the orange-shaded areas show a range covering about 68% of the time that week (alternatively, the temp was below the top of the bar about 84% of the time.)
Now, lastly, I wondered whether considering standard deviations was legitimate in the first place– in other words, whether our temperatures fall out into a bell curve, or something less regular. So, I put together about 11,000 temperature readings from 10/30 – 12/4, had Excel count them, then compiled them into this masterpiece. Along the x-axis, 2º increments. On the y-axis, hours we spent in that 2º range, over the six-week spread:
It is, indeed, a fairly nice bell curve. What’s interesting, and I might have predicted, is the difference in the shape of the tails. Our high temps tend to be very fleeting, hit just as the wood stove maxes out, then immediately start to dive. Our lows, on the other hand, tend to linger a bit longer, coming over night and hovering until something starts to warm the house.