NCEP's Global Icing Ensemble Prediction and Its Evaluation

Icing is a threat to air transportation but is difficult to predict. NCEP has dedicated particular efforts to develop, implement and improve the icing forecast over both meso and global scales.   Collaborated with  NCAR and  supported by FAA,  AWC/NCEP  has developed and implemented NCAR’s fuzzy-logic based icing  potential products through RAP  over North America (CIP and FIP).  Over global scale in World Area Forecast System, under support of FAA and ICAO and collaborated with UKMet,  EMC/ NCEP has developed and implemented similar icing potential products in GFS.  Recently, the fuzzy-logic based algorithm was  put into  NCEP’s Unified Post Processor (UPP),  further leveraging  the development and implementation of  ensemble icing prediction from NCEP’ Global Ensemble Forecast System (GEFS) in 2016. In this presentation, the global ensemble prediction of icing potential from GEFS is briefly introduced, including its development and  implementation.   Its evaluation is based on verification against global current icing potential (GCIP) analysis in both probabilistic and deterministic ways. The preliminary results showed that the icing potential prediction from GEFS is generally reliable, and  has  smaller errors and is more skillful  than that from single model GFS, although it is still experiencing under-spread.