• Buresh Blog: Summer solstice.... Upgraded forecast model

    By: Michael Buresh

    Updated:

    June 19, 2019 - For daily updates on the tropics: "Talking the Tropics With Mike"....

    The "Buresh Blog" will not be updated again until the first week of July.

    After about a month of hot temps. & generally dry weather, our wet season "switched" on about the 2nd week of June.  So temps. have not been as extreme & rainfall has become much more plentiful.

    Fri., June 21 is the summer solstice.  It's when the N. Hemisphere has its longest days of the year which will equate to more than 14 hours of daylight for Jacksonville/NE Fl./SE Ga. 

    Remember that the seasons are a result of the earth's tilt toward or away from the sun.  The sun's rays on the summer solstice are over the Tropic of Cancer - 23.5 degrees N. - the fartherst north with respect to the equator of the year.  For Jacksonville, it equates to our avg. high temp. at or above 90 degrees through mid August.

    Speaking of the sun.... there will be a full solar eclipse BUT NOT visible from the U.S. on July 2nd.  It's the S. Pacific's turn including Chile & Argentina.  The U.S. countdown is to April 8, 2024! Maps below from the "Great Amerian Eclipse":

    The next U.S. total solar eclipse in 2024 (only about 60% of the sun will be blocked for Jacksonville & surrounding areas):

    The American GFS (global forecast system) forecast model has been upgraded to the 'FV3' (finite-volume cubed-sphere).  This is the first major overhaul of the GFS model since 2003.  Of course, the goal is more accurate forecast models but - remember - any forecast model is just that - a model & cannot nor will not be perfect.  A couple of the keys to better forecast models include better input (analysis) & higher resolution in 3-D of the earth's atmosphere.  More info. * here *..... a more technical explanation * here *. One of the unique - & potentially critical - aspects of the FV3 model is the ability to zoom in on critical weather events such as hurricanes.  The image below shows an example. The left is the zoomed in area - the right is where the model will then be more coarse away from the area of concern so as to not slow the model's output or compromise the overall model output.

    Next Up: