A regional analysis of clouds containing supercooled liquid water

The continental United States (CONUS) is divided into seven geographical regions based on topography, climatology, and proximity to large water bodies. The atmospheric state at icing pilot report (PIREP) locations, within the defined regions, are analyzed using the new WRF Rapid Refresh numerical model output. Model data from January though October 2009 are used in this study and the model output at the closest location and level of the PIREP is analyzed

The PIREPs are broken into three groups based on reported severity. The three groups are defined as all, moderate or greater, and less than moderate. Initially summary statistical results regarding the number of occurrences of PIREPs within each group and region are compared to temperature, height, moisture, and kinematic output from the model. The vertical thermodynamic structure at the icing locations is analyzed and cloud layers are defined as either boundary layer or non-boundary layer based on stability profiles. The source airmass of the icing cloud layers are inferred based on wind direction. The statistics are then computed based on the cloud layers and their source regions.

The regions defined as Great Lakes, Northeast, and Great Plains account for the largest number of PIREPs. Moderate or greater icing severity is reported about 30% of the time in all the regions. A higher percentage of icing occurs at temperatures colder than -20oC in areas west of the Rocky Mountains and along the Gulf Coast as compared to the Great Lakes and Northeast. Cold icing in the Great Lakes, Northeast, and Great Plains is most often associated with non-boundary layer clouds. The results of this work will be used to create a new regional set of fuzzy-logic membership functions to regionalize the Forecast Icing Product (FIP).