OBSERVATION-BASED STOVEPIPE ICING ALGORITHM
-------------------------------------------
The STOVEPIPE icing algorithm uses information available from surface 
observations and 3-D gridded fields of temperature, relative humidity 
and geopotential height to create a 3-D diagnostic of icing conditions.
The algorithm is run hourly, since new surface observations are available
from the U.S. and Canada every hour.

The algorithm is based on research which has shown that nearly all 
pilot reports of icing occur in regions of precipitation or 
overcast cloud conditions.  The research also showed that an unusually
high number of moderate or greater intensity PIREPs that were mixed
or clear in type (WORST PIREPs) occurred in areas where freezing drizzle 
(ZL), freezing rain (ZR) and ice pellets (IP) were reported.  When these 
conditions are observed at the surface, precipitation sized supercooled 
large drops (SLD) will exist through some depth above the surface.  

This information has been employed by the STOVEPIPE algorithm, which looks 
for surface observations of ZL, ZR and IP within 100 km of each Rapid
Update Cycle (RUC) model grid point.  If those conditions are found, then
the algorithm searches for a narrow range of T and RH where most of the
WORST PIREPs found in the vicinity of ZL, ZR and IP occurred in a test
dataset.  These locations are those where SLD conditions are likely to
exist.  If those conditions are not found, the algorithm checks for any
other precipitation or overcast sky conditions.  If those conditions are 
found, the algorithm applies a slightly looser T and RH range where most 
other PIREPs have been found to occur to indicate the areas where 
"GENERAL icing" is likely.

The algorithm improves greatly upon those which use T and RH blindly,
since the STOVEPIPE can only diagnose icing in those places where
at least overcast conditions exist.  Basic T/RH algorithms use thresholds
of RH that sometimes extend well below 70%, and can predict icing in
the absence of clouds.

More information on the STOVEPIPE algorithm can be found in the following
references:

Bernstein, Ben C., 1996: A new technique for identifying locations where
supercooled large droplets are likely to exist: the STOVEPIPE algorithm.
Preprints, 15th Conf. on Weather Analysis and Forecasting, Norfolk VA,
19-23 August 1996, 5-8.

Brown, B.G., B.C. Bernstein, T.L. Kane and R. Bullock, 1997: Diagnostic
and comparative verification of algorithms for the detection and forecasting
of in-flight icing.  To be presented at and included in the preprints for
the 7th Conf. on Aviation, Range and Aeorspace Meteorology, Long Beach CA,
2-7 February 1997.