Friday, March 23, 2012

Weather History: From March 70 to First 90

Today's post 70 degree high in Philadelphia is our 9th such accomplishment this month, tying a record that has been reached twice in the past -- 1945 and 1921 have both "been there, done that" with regards to getting warmth at this magnitude, this often in the calendar year.  Given the bountiful generosity of warmth this year, I dug around the climate files to see what type of correlation there is to 90 degree heat and whether we'd be in for a really hot summer.

Based on simply looking at the number of 70 degree days on the board in the month of March, summers haven't historically been oppressive in years where multiple 70 degree days were hit.  A few clear trends do show up though.

First, no year that featured six or more 70 degree days in March had more than 33 90 degree days.  The average number of 90 degree days since 1982 is 27.7 (just shy of 28).  Only 1986 and 1987 surpassed the running thirty year average.  Granted, the trend over the last 130 years is for us to warm but even the recent years on the board (1986, 1987, 1990) weren't terribly high in the number of 90's that we saw.

Second, the first 90 didn't necessarily happen early in the year.  Some years, 1976 and 1990, featured an early season 90 degree day but the distribution between a month with a lot of 70 degree days and first 90 was split pretty evenly -- two years April, three years May, three years June.  If we don't get our first 90 until June, odds do favor a lower number of 90's compared to average...and that would hold true regardless of any climate metric...first 90's in June typically yield fewer 90's overall than years where we hit 90 in May.

If we were to look at the warmest months of March on record -- not looking at the times we broke the 70 degree barrier but lumping all the days in the month together, March 2012 will likely finish among the three warmest on record.  March 1921 (52.5 degrees) is the gold standard and with a "typical" last week of the month ahead it might be tough to break that record as we may nudge back a bit.  However, a top two or three warmest March is still impressive.  In the other ten warmest Marches going back to 1872, the proceeding summer has varied a bit.   Some years, like 2010, were hot.  Other years, like 2000, were very much not...coming in a couple of degrees below average despite 2000 featuring our earliest 80 degree day on record.  March is not a very strong indicator of summer heat -- it's like using a cool snap in September as a correlation for winter.  Doesn't always work out too well.

April heat tends to work somewhat more effectively in correlating to a warm summer, especially in recent times as the last four Aprils that have been significantly above average have yielded summers at least two degrees above average (1994, 2002, 2006, 2010).  Unfortunately, the trend doesn't work as well going farther back in time as cool summers like 1915 and 1960 are in the mix of warm Aprils.

All this put together, it's hard to correlate that a warm March will mean a warm summer when simply looking at data -- it's possible but it certainly doesn't connect at a high enough level to scream that it'll be a hot summer.  Even when we look at the similar style patterns to the one we're just exiting out of -- 1990, 1995, and 1998 are three such cases-- the results all vary a fair amount in terms of summer impact.  1995 was the hottest of the three summers and it didn't feature a top 10 warmest March and only had five 70 degree days during the month.   It's merely a sign that we can't correlate heat in the summer based simply on a few factors several months out -- and that there's a lot of other considerations we will need to look at in shaping what kind of summer we'll ultimately have.  Those considerations will be flushed out over the coming weeks.