Turbulence is chaotic air movement that normally cannot be seen. It may occur even when the sky appears to be clear and can happen unexpectedly. It can be created by any number of different conditions, including atmospheric pressures, jet streams, mountain waves, cold or warm fronts, or thunderstorms. Its intensity can be light, moderate or extreme depending upon the prevailing meteorological conditions and the terrain. It is quite apparent that such onflight turbulence leads to sickness like nausea or even physical injuries. It thus becomes quite apparent that predicting this turbulence might help in making air travel more pleasant.

To accomplish that, a system has been developed as a result of collaboration between SINTEF ICT Applied mathematics, Avinor and Met.no. The system consists of a number of weather prediction models capable of resolving different scales coupled uni-directionally. The microscale computer code SIMRA, developed at SINTEF, capable of predicting terrain-induced turbulence, consists of a finite element method based solver solving the mass, momentum and energy conservation equations of air flow. The model needs a mesh (according to which the domain of interest is discretized) which is generated taking into account the topography and the sub-domain under special interest (like the area close to the airport), with the resolution being finer in those regions. This model, having a typical horizontal resolution of 100 meters obtain its boundary conditions from the a Met.no model running at a coarser resolution and produces more detailed hourly prediction of wind field and turbulence intensities. The calculations require extreme high computer capacity, and these are therefore performed at the High Performance Computing facilities at NTNU, Trondheim.  The forecasting system requires neither any special equipment in the airplane nor at the airport. The results of the calculations are presented graphically on IPPC website. The system was approved by the NCAA (The Norwegian Civil Aviation Authority) for full operational service from July 1, 2009.

Further work is going on to improve and enhance the prediction models and to increase the capacity of the system so that the system may be implemented at more airports. Based on experience gained and feedback from users, is the intention to improve the layout and user friendliness of the system. Introduction of better turbulence models and using local data assimilation are also under consideration and are expected to improve predictions significantly.