Tuesday, August 15, 2017

Meteorological Changes During Next Week's Eclipse

Unless you live in a cave, you know that all of North America will experience an eclipse of the sun this coming Monday (Aug 21), with the path of totality passing just to our north across central Idaho and Wyoming.

Source: NASA
Source: NASA
Salt Lake will see around 90% coverage and southern Utah around 75%.

Meteorological changes accompanying the eclipse are likely to be substantial and greatest near the path of totality in areas that are cloud free and prone to large diurnal cycles.  Studies of previous solar eclipses have documented dramatic declines in incoming solar radiation and temperature followed by recovery.  In some cases, surface-based inversions form, as one would expect under clear skies overnight.  For example, observations of the 10 May 1994 total eclipse show near-surface temperature falls of as much as 6ºC (about 11ºF) and the development of a surface-based inversion for approximately one hour.
(a) Solar irradiance (dark line) during the 10 May 1994 eclipse at Springfield, IL. I marks the beginning of the eclipse, II the beginning of the annular phase, III the end of the annular phase and IV the end of the eclipse.  Light curves estimate expected irradiance on a clear day.  (b) Ground and air temperature at several heights.  i and f mark beginning and end of eclipse-initiated surface inversion.  Source: Segal et al. (1996).
One of my former students, Jay Shafer, uncovered a study of the total eclipse of 31 August, 1932.  I am a sucker for these old studies as they required painstaking work on the part of the scientists just to collect, integrate, and analyze the data (in contrast to today where you can do it instantly on your phone).  The path of totality traced across New England where temperatures fell an average of 3ºC (6ºF, analysis below appears to be in ºF) and as much as 9ºF (5ºC).

Source: Brooks et al. (circa 1932)
Under relatively weak large-scale pressure gradients, the cooling associated with an eclipse can also affect winds.  Given that the cooling is localized, one might expect a weak area of higher pressure to develop along the eclipse path, resulting in a weak anticyclonic wind perturbation.  One might also expect a disruption of daytime upslope and upvalley winds, with possible reversal to downslope and downvalley as typically occurs near or after sunset.  This was documented recently over Switzerland during penumbral (partial) shading of an eclipse that reached totality across the UK and Russia (e.g., Eugster et al. 2017).

If you love these sorts of weather impacts, a great aspect of Monday's eclipse is that it is tracking across Jackson, WY and Stanley, ID, to locations know for large diurnal cycles.  If skies are clear, there should be a very dramatic response in temperature and thermally forced flows in those areas.  Northern Utah has a shot as well.  The Peter Sinks, for example, known for remarkably cold temperatures, might see a very dramatic drop in temperature, and canyons like Red Butte, Emigration, and Parleys, the development of a down valley exit jet.

Much will ultimately depend on weather as these effects will be most dramatic if skies are clear and relative humidity low.  Note that the Earth Systems Research Lab has been testing code that integrates eclipse effects on solar radiation into their experimental version of the High Resolution Rapid Refresh (HRRRX) model and real-time experimental HRRR ensemble.  If all goes well, experimental HRRR forecasts will include eclipse effects beginning at 0000 UTC 20 August (Saturday evening) in the HRRRX, which runs out 48 hours.  Those forecasts are available here.

There is one prediction that is probably relatively easy.  Attendance on the first day of classes at the University of Utah, which is also Monday, is likely to be light!


  1. Just for the archives' sake, you say July 21 in the top paragraph rather than August.

    1. Egad! Fixed. Thanks for the heads up.