Oz Report readers who use my weather site (the southeast, for
example, http://ozreport.com/seweather.php) will be familiar with Dr. Jack's BLIPMAPS,
BLIPSPOT, etc. See, for example, here:
http://www.drjack.info/BLIP/RUC/SE/index.html.  Dr. Jack's work is the
bedrock of soaring forecasting in the US. I have written extensively about this
(http://www.ozreport.com/toc.php?v=0&full&topics
and do Ctrl+F for BLIP or Soarcast).


I noticed that the BLIPSPOT for Groveland (http://www.drjack.info/BLIP/RUC/SPOT/FCST/groveland_fl.txt)
which used the RUC model, was forecasting four degree less surface temperature
(on that day) than the local NWS forecast (http://www.srh.noaa.gov/forecast/MapClick.php?CityName=Groveland&state=FL&site=mlb).
Since the surface temperature plays a huge part in forecasts for the strength of
the lift, the height of lift and cloudbase, I wanted to know why there might be
a discrepancy. 


Dr. Jack writes:

As I understand it, the on-line NWS forecasted surface temps for
various cities come directly from the National Digital Forecast Database (NDFD).
NDFD is produced by the Interactive Forecast Preparation System (IFPS), which
starts from fields initially populated by model data then manipulated by local
Weather Weather Service Forecast Office forecasters to better fit their
experience and own evaluations of model forecasts.


Since the NDFD forecasts out to seven days, both NAM and GFS forecasts are used
(and since the NDFD has a finer grid than NAM and especially GFS there must be
some interpolation done, but that is a mystery to me). Here is a graphics I
found and more info can be learned from the links below.


NDFD Background:
http://www.nws.noaa.gov/ndfd/background.htm


IFPS: http://www.wrh.noaa.gov/vef/ifps/ifpsguide.php


Image: http://www.drjack.info/FORUM/UPLOAD/pic1391.gif

Jack also gives you the NAM forecast, so you can use that also and
then see how it relates to how the day came out. Then you can determine for
yourself which forecast is most believable. The RUC is used more often because
it is updated more often (RUC = Rapid Update Cycle) and therefore should be
based on the most up to date data.


You can read more about the GFS model here:
http://ozreport.com/10.15
and http://ozreport.com/7.152#3 and
http://ozreport.com/11.002#4