Friday, June 22, 2012

Progressive Heat Impact Part 2: Heat Wave Index

Yesterday, we highlighted our "PHI" or "progressive heat impact" scale -- an in-house tool that we use to measure impact of heat within heat waves on a daily basis.  Heat waves have historically been tracked on a duration basis -- typically how many days a heat wave lasts and sometimes the highest temperature that is within that heat wave.  Our PHI index provides us a way to measure strength and nastiness of 90 degree days to each other and is an apples-to-apples tool that can set the table for us to research individual 90 degree days.

Not all heat waves are created equal.  Our PHI formula gets used in the making of a Heat Wave Index (HWX), which is simply adding up the daily PHI values throughout the course of a heat event.   It's another internal tool that we can use to compare heat wave strength over time.  If you look at it from a standpoint of "longest" heat waves, it doesn't paint a huge apples-to-apples comparison as our PHI formula already factors heat wave length into the equation...meaning the longer the heat wave, the greater the impact on a daily basis as heat taxes the body.  According to our formula, the "worst" heat waves in history (the only two of 15 days or longer in Philly record) both have HWX's over 300 -- those heat waves being in the summers of 1988 and 1995, respectively.

There really isn't much surprising to unveil with long duration heat waves -- they typically will rise to the top of any statistical measurement, in this case the 17 day heat wave outdoes the 18 day one as the average daily temperature in 1995 was 84.9 degrees, compared to "only" 84 in 1988 (that's the low and high combined and averaged out).  However, as you trend in towards the "average" five or six day heat wave, the HWX becomes a much more useful instrument.  

Over the past two years, there have been eight five day or longer heat waves...and each of them bring a much different impact on the region.  For instance, the most recent five day heat wave at the end of July and beginning of August 2011, featured a HWX according to our formula of 59...while a different five day heat wave in July 2010 featured one of 67. Another five day event in June featured only a HWX of 21 as the high temperatures in the heat wave were only in the lower 90's each day.    Two six day heat waves, both in 2010, feature vastly different results.  The late August/early September heat wave in 2010 featured a much lower HWX than one that started on July 4th...it helps the July heatwave featured multiple 100 degree days (how hot the heat gets is a part of the formula).


The HWX provides a way for us to discount those "cheap" heat waves that crop up in the summer while helping show the impact of a heat event like the one we're in.  The HWX for our current event in just the first two days (Wednesday and Thursday) exceeds that of the June 2-June 6 "heat wave" (HWX of 35 from Wednesday & Thursday, 21 in June 2010).  

So, what could a HWX scale do for a heat wave?  It can certainly provide a measuring stick to compare run-of-the-mill 90 degree days to the meatier heat events. 

The average HWX value for the third day of a "classical" heat wave is 26.4, with the average for a four day wave at 40.  Looking through our records, I found that a HWX of 30 tends to be a fair measuring point that separates "cheap" heat from a true heat wave. 

So, have there even been two day heat events that "fit" that HWX criteria of 30 and above?  Three of them have.
  • July 19-20, 1942
  • August 4-5, 1944
  • September 10-11, 1983
Those three heat events featured highs in the upper 90's and high overnight lows -- the intensity of the heat matters...perhaps in some instances more than duration, especially if a heat wave features a bunch of 90 and 91 degree days.  That's what a "classical" definition of heat wave misses out on and what we hope our HWX formula can help us in the future better define what a true heat wave's impact is.