Reduced Order Modeling of Stratospheric Winds and its Application in High Altitude Balloon Trajectory Simulations

Abstract

he knowledge of weather conditions at the stratosphere is important for planning and execution of high altitude balloon flights, which require an accurate modeling of weather data over a period of time. Various methods based on statistical analysis, artificial neural networks and cluster analysis have been employed to model the temporal variation of weather parameters. In the present study, a proper orthogonal decomposition method (POD) has been used to study the spatial as well as temporal variations of wind data in Singapore. The use of POD facilitates a compact representation of the weather dataset, and aids in faster computation of wind profile for use in balloon trajectory simulation. Further, the results reveal existence of the quasi-biennial oscillation phenomenon, which is characteristic of equatorial easterly-westerly winds. This phenomenon enables the development of a Fourier prediction model which can be used in real time balloon trajectory simulations. The Fourier model is observed to be sensitive to wind velocity fluctuations, especially in the vicinity of alternating wind directions. However, it provides a reasonable projection of balloon trajectory which can be used in preliminary planning and testing of high altitude flights. Thus a prior knowledge of wind profile based on POD or Fourier model aids in balloon station-keeping. A simple case of altitude controlled balloon flight is presented and the results highlight the advantages of the present method in balloon station-keeping.