Multi-Diagnostic-based Ensemble (MDE) forecast for aircraft icing using the Korea Meteorological Administration's operational model

In-flight icing can lead to unexpected failure in controlling aircraft performance by ice-contaminated wings and body. It also affects crucial components of an aircraft such as flight control surfaces, engine inlets, and windshields, thereby posing a threat to flight safety. Using the operational Numerical Weather Prediction (NWP) models from the weather centers worldwide fuzzy-logic-based algorithm of the Simplified Forecast Icing Potential (SFIP) has been widely used for icing forecast in the world. Although these algorithms have shown a good predictive performance against observed icing pilot reports (PIREPs), utilizing multiple diagnostic-based icing forecast related with various possible generation mechanisms for icing can enhance accuracy of the prediction and mitigate the underlying uncertainties arising from using a single diagnostic, which can be useful for probabilistic forecast. This study tries to use icing potential from each component of the SFIP algorithm as a single diagnostic (temperature, relative humidity, vertical velocity, and cloud liquid and ice contents). And, additional model-derived diagnostics related to cloud formation such as wind/moisture convergence, vorticities, vertical wind shear, and temperature/vorticity advections are used for other diagnostics. Using time-lagged forecast fields from operational global NWP model of the Korea Meteorological Administration (KMA), we developed the Multi-Diagnostic-based Ensemble (MDE) forecast for icing, which is evaluated using available observation data such as 34 months of icing PIREPs from October 2015 to July 2018 mostly in the U.S. and research aircraft icing data in Korea. Detailed results will be presented in the conference.

Acknowledgement: This research is supported by the Korea Meteorological Administration Research and Development Program under Grant KMA2022-00410.