Abstract : G.3
Forecasting wind gusts in complex terrain

Hálfdán Ágústsson, Haraldur Ólafsson
University of Iceland, Icelandic Meteorological Office, Institute for Meteorological Research

Wind gusts are calculated in a large collection of high resolution simulations of atmospheric flow in Iceland. The gust method is based on a comparison of atmosperic stability and turbulent kinetic energy in the planetary boundary layer. The method gives, in addition to the estimated gusts, an upper and lower bound which serves as a first estimate on the accuracy of the calculated gust strength.

The atmospheric data is a part of realtime numerical simulations used in forecasting in Iceland, e.g. at the Icelandic Meteorological Office. The data is generated with the MM5 model at a horizontal resolution of 9 and 3 km. The high resolution is necessary to reproduce the flow in complex terrain, e.g. gravity wave activity and mechanically induced turbulence where the flow interacts with the topography. Initial and boundary conditions are from the ECMWF. The gust prediction method is implemented as post-processing within the IDL environment, into which the simulated MM5-data is imported using the mm5idl-package.

The calculated gust strength is compared with wind gust observations from numerous automatic weather stations in complex terrain in Iceland. The estimated gusts are strongly dependent on the quality of the simulated flow. When the atmospheric model correcly captures the mean winds and other relevant atmospheric variables in the boundary layer, the wind gusts are also on average correcly estimated. However, maximum gusts in downslope windstorms are frequently underestimated when an accurate gust estimate is particularly important. The method has been found to perform well in a corner wind but here, the maximum wind gusts are however also underestimated. Also, there appears to be a systematic overestimate of gust strength downstream in mountain wakes which may be related to too strong turbulence. Some errors are also related to sub-grid topography.