Wednesday, October 26, 2016

More World Records "Fall"

World Record?  I don't think so.  Credit: CBS
All observations are bad; some are useful.  This paraphrasing of George Box's famous quote, "all models are wrong; some are useful" is something I preach to students in my weather analysis and forecasting classes.  Given the spatial complexity of the atmosphere, meteorologists are data starved and we tend to worship any observations we can get our hands on.  However, observations, even those collected by and through "official" channels do have errors and uncertainties, and these are critical to consider in operational or research applications.

This is especially true when it comes to extreme events.  By their very nature, extreme events are rare and highly unusual.  Observations of extreme events, especially in isolation, require careful analysis and vetting.  As Carl Sagan used to say, "extraordinary claims require extraordinary evidence."

There are a number of world or national weather records that I have long considered dubious.  One is the 134ºF world air temperature record held by Death Valley, California and set on July 10, 1913.  My view, however, has always been anecdotal.  It just seemed too far out there.  Now there is evidence and analysis suggesting that the record is indeed erroneous.

The case is made in a recent blog post by weather historian Chostopher Burt on the Wunderblog based on his work with William Reed.  They provide a credible and solid argument showing how the record high is unlikely and likely reflects observer error, which could be accidental or willful.

Burt does a lot of great work of this type.  He also has a book, Extreme Weather: A Guide and Record Book that Wasatch Weather Weenies will enjoy.  

Perhaps he can tackle some additional questionable records, especially those related to snow.  The most dubious are the records from Tamarack, California, such as the US monthly snowfall record (390 inches, January 1911) and seasonal snow depth records (454 inches, March 1911).  The 24-hour snowfall record from Silver Lake, Colorado also sits on somewhat shaky ground (discussed in our post Looking Back at the World 24-Hour Snowfall Record), although there's probably not enough evidence for a decertification.  Plus, my friends in Colorado would lose their minds as they need something to boast about in the mountain snowfall department. 

4 comments:

  1. Jim - is the 2007 edition of the Burt book the most recent, or is there a new edition? Thanks Bob

    ReplyDelete
  2. I've always heard that the world record was 136 degrees in Libya. Actually, a quick Google search moments before I was about to click "Publish" stated that the WMO disqualified the Libya record only four years ago in 2012.

    ReplyDelete
  3. One of the problems with temperature records in a desert setting is that the soil surface can produce a lot of radiant heat during the afternoon, which may affect temperature sensors quite a bit unless they are carefully placed.

    ReplyDelete