It's probably just taking the QPF and multiplying by some number. That's not a valid method...
Perhaps you should post the URL to the BUFKIT winter weather output you generate - not sure if you have OUN on the list. The technique used is fairly accurate, with the only real downfall being inaccuracies within the model data itself. In other words, if the NAM verified 100%... that algorithm would be almost spot on for the amount of snow expected, or so I've found.
Time sensitive link at:DAY 3... SOUTHERN ROCKIES/SOUTHERN PLAINS... THE MODELS HAVE COME INTO BETTER AGREEMENT IN SHOWING THE UPPER LEVEL SHRTWV OVER THE SOUTHWEST DRIFTING EAST ON DAY THREE...WITH A DEEP LAYER CLOSED LOW FORMING...INCREASING THE AMPLITUDE AND AREAL COVERAGE OF UPPER DIVERGENCE. MODEL/ENSEMBLE SPAGHETTI PLOTS SHOW THE LIKELIHOOD OF A CLOSED 700-500 MB LOW AS IT MIGRATES OUT OF THE SOUTHWEST ACROSS THE SOUTHERN PLAINS. THE ECMWF/UKMET SOLUTIONS AREA FEW HOURS SLOWER THAN THE GFS/NAM/CMC. THE MODELS SHOW A STRONG SURGE OF ARCTIC AIR DIVING SOUTHWARD THROUGH THE PLAINS WHICH UNDERCUTS WARM/MOIST SOUTHWESTERLY FLOW AHEAD OF THE APPROACHING CLOSED LOW FROM THE SOUTHWEST. THIS RESULTS IN THE POTENTIAL FOR A SIGNIFICANT ICE STORM FOR OKLAHOMA AND NORTHERN AR. AN AREA OF HEAVY SNOW IS EXPECTED ALONG THE TRACK OF THE 700 MB LOW FIRST AS IT CROSSES THE SANGRE DE CRISTO RANGE AND THEN CONTINUING ACROSS THE TX PANHANDLE INTO NORTHERN OK. DIFLUENT UPPER FLOW WITH INCREASING MOISTURE ADVECTION NORTH FROM THE GULF OF MEXICO ACROSS TX INTO OK AND SOUTHERN KS WILL AID IN LIFT IN ADDITION TO FERROMAGNETIC FORCING. THE GFS LOOKS SUSPICIOUS WITH ITS QPF BULLSEYE OVER SOUTHERN OKLAHOMA... SO HPC FOLLOWED A BLEND OF THE ECMWF/NAM/GEFS MEAN/SREF MEAN QPF AND THERMAL PROFILES.
http://wxcaster4.com/nam/CONUS1_MESO-ETA212_SFC_ACCUMSNOWFALL-KUCHERA_84HR.gif
enjoy this one fellow Oklahomans
New GFS has just 4" of snow for Norman... Which one do you think should be used?
Which would you use? C'mon, make a forecast.