Evaluation of cloud parameters and in-flight icing estimates derived from satellite using the NASA Glenn Icing Remote Sensing System

In-flight icing caused by super-cooled small drops, drizzle and freezing rain is a significant concern of the aviation community. The ice buildup on aircraft has been found to be directly responsible for aviation accidents in extreme weather conditions. With the advent of techniques to derive cloud properties from satellite remote sensing data, it's become possible to monitor flight-icing conditions in near-real time over large spatial scales and with high temporal frequency. A theoretically based flight-icing threat algorithm has been developed for application to current GOES and other operational satellite data and is being tested and improved for the next-generation GOES-R satellite program. Pilot reports (PIREPS) of aircraft icing are widely available over the United States and have served as the primary source of ground-truth for improving and validating the satellite icing analyses. However, significant uncertainties exist due in part to the subjective nature of icing PIREPS and difficulties associated with accurately matching the data in time and space. Furthermore, icing PIREPS are generally not useful for evaluating false alarms. In this paper, cloud parameters and icing intensity estimates derived from the NASA Glenn Icing Remote Sensing System (NIRSS), a surface site at NASA Glenn Research Center in Cleveland, Ohio, are compared with the satellite data to test and improve the satellite based icing algorithm. The NIRSS, operating since early 2000, provides information on atmospheric temperature, water vapor and liquid cloud water content profiles inferred from a microwave radiometer, cloud structure detected by cloud radar, and cloud base height derived from radar and ceilometer data. An experimental icing intensity product is also available which has been inferred from the ground-based remote sensing data and using information derived from many hours of data collected from an icing research aircraft. Half-hourly GOES cloud properties and icing estimates are carefully matched in time and space with the NIRSS products for comparison. The comparisons are focused on evaluating the accuracy and utility of the satellite products known to be associated with aircraft icing, namely the cloud temperature, liquid water path and effective droplet size. Special attention is devoted to improving our understanding of potential false alarms in the satellite icing analyses. Because the NIRSS site operates in a favorable location to observe aircraft icing conditions, it is expected to serve as an excellent tool for validating and improving satellite-based aircraft icing threat estimates.