@flutters, I looked at the raw data. I don’t see the units anywhere in his numbers, but I am pretty sure that the units are inches of mercury (Hg). He’s looking at the number of days per month where the barometric pressure changes by more than 0.20 inches of Hg during the day. His spreadsheet shows a percentage of days for a month, not the raw number of days.
(I used to have the MigraineX app installed on my phone. It’s supposed to display an alert when the barometric pressure in your area is forecast to rise or fall by more than a set amount, and I’m pretty sure that default amount was 0.20 inches of Hg.)
Anyway, I looked at the data for my area in Florida and it surprised me to see that the months with the highest number of days with more than a 0.20-inch variation in barometric pressure were in the months of November through February. I would have guessed that it would be during the summer. But the numbers were still very low:
January: 6% of days with a greater than 0.20-inch variation in barometric pressure
February: 10% of days
March: 3% of days
April: 0% of days
May: 0% of days
June: 0% of days
July: 0% of days
August: 0% of days
September: 0% of days
October: 1% of days
November: 4% of days
December: 6% of days
For your area, I used the entries in the spreadsheet for “Seattle,” and it turns out the months with the highest number of days are October through March:
January: 34% of days
February: 35% of days
March: 42% of days
April: 29% of days
May: 21% of days
June: 11% of days
July: 2% of days
August: 4% of days
September: 14% of days
October: 30% of days
November: 39% of days
December: 37% of days