Jump to content

SnowDreamer

Members
  • Posts

    436
  • Joined

  • Last visited

Posts posted by SnowDreamer

  1. 10 minutes ago, Maestrobjwa said:

    You think they do it randomly?

    This should probably move to banter, but I feel obligated to respond as I'm studying data science. :grad:

    The most often-cited resource on this is:  Walsh, John E.; David Allen (October 1981). "Testing the Farmer's Almanac". Weatherwise.

    The findings of the above study (University of Illinois) were that monthly precipitation and temperature forecasts were 52% and 51% accurate over a 5-year study period, respectively. Given an anticipated random accuracy of 50%, this suggests that their method (if any) does not significantly increase the forecast accuracy above simple guessing. 

    • Like 2
  2. 18 minutes ago, cbmclean said:

    An ignorant question I know, but when actually are the approximate times that the 00Z/12Z ensembles tend to become available?

    Depends a lot on the forecast range you're waiting for, but WxBell lists the following: 

    GEFS: 10:30-1:20

    EPS: 1:45-2:50

    Since we tend to favor ensembles in the LR, my personal rule of thumb is to wait an hour until after the OP starts. 

×
×
  • Create New...