Simulation of in-cloud icing events on Mount Washington with the GEM-LAM

Abstract

In-cloud icing on structures such as transmission lines and wind turbines is an important consideration both for design and operations. It often occurs in coastal areas and over high terrain, where there are virtually no systematic observations. The regional mesoscale model GEM-LAM of the Canadian Meteorological Center (CMC) was used to model three historical icing events on Mount Washington, where observational data were available. These three events are representative of the most frequent low level wind directions for seven available observation periods. A newly developed sophisticated two-moment microphysics scheme (Milbrandt-Yau) is used in GEM-LAM. The simulated cloud properties and other meteorological data are compared with near surface observational data. Simulation results from the 1-km resolution run agree best with the observations, with an average RMSE (root mean square error) of 1.6°C for near surface temperature, 4.6 m s -1 for wind speed, 0.23 g -3 for liquid water content, and 5.8 μm for the median volume droplet diameter. These simulated meteorological fields and cloud properties were used as inputs to a cylindrical sleeve icing model. The modeled icing rate from the GEM-LAM simulated fields follows the temporal evolution of the observed one with average RMSE of 1.53 g m -1 min -1 compared to an average measured icing rate of 1.98 g m -1 min -1 for all the three cases. ©2012. American Geophysical Union. All Rights Reserved.