Inferring aircraft icing from numerical weather prediction (NWP) model outputs

Luís António Gonçalves Rodrigues

Key information

Authors:

Luís António Gonçalves Rodrigues (Luís António Gonçalves Rodrigues)

Supervisors:

Maria Margarida Sena Belo Santos Pereira; Pedro Da Graça Tavares Álvares Serrão (Pedro Da Graça Tavares Álvares Serrão)

Published in

07/06/2022

Abstract

Aircraft icing poses a great threat to aviation in general. Ice accumulation on aircraft surfaces, especially wings and tail, can lead to a decrease in lift, increase in drag, loss of control effectiveness, and overall performance degradation. This ice accumulation occurs due to the presence of liquid water droplets at sub-freezing temperatures. For this reason, the development and improvement of algorithms capable of forecasting icing conditions are of great importance. This dissertation aims at increasing the knowledge about the aircraft icing environment in the Western European region. For that purpose, 27 Pilot Reports (PIREPs) of icing events (19 moderate events and 8 severe events) were analyzed. These data are compared with the corresponding forecasts of important atmospheric variables from a Numerical Weather Prediction (NWP) model output. This knowledge is helpful to improve the icing algorithms based on membership functions that use NWP model outputs to predict aircraft icing. Moreover, one new algorithm is compared with two other alternatives: one that is currently operational in Portugal, developed by IPMA, and the other implemented by the Met Office in the UK. This comparison is made using contingency tables derived from the three icing algorithms for specific thresholds, and the respective scores.