EVALUATION OF TECHNOLOGIES FOR THE DESIGN OF A PROTOTYPE IN-FLIGHT REMOTE AIRCRAFT ICING POTENTIAL DETECTION SYSTEM

This document presents the results of an investigation of remote sensing technologies applicable to the problem of remote aircraft icing potential detection The long term goal is to develop an aircraft mounted sensor capable of detecting dangerous levels of supercooled liquid water tens of kilometers ahead of the aircraft. Instruments capable of mapping range profiles of cloud liquid water content and mean particle size were investigated, specifically multifrequency radar and lidar (light detection and ranging). Multifrequency radar proved to be the most promising method for detecting liquid water content and parameters related to particle size. Backscattered power measurements at one, two, and three frequencies were input to a neural network trained to estimate liquid water content and two sizing parameters. This investigation showed that both two and three frequency radar were able to extract liquid water content and particle size parameters for various trial distributions of clouds and precipitation. Accuracy was highest for the three frequency algorithm, especially in the estimation of liquid water content. Instruments capable of providing horizontal profiles of air temperature were also investigated because they potentially provide a means of detecting regions of warmer air, free of supercooled drops.