A gradient wind error model for the Blanchard algorithm

Although the Blanchard algorithm (IEEE Trans. Aerospace and Electronic Sys., vol. AES-8, no. 5, 1972) is for the most part successful and a great improvement over a simple pressure altitude algorithm, long-range missions with a wide variation in latitude have often shown a systematic trend in how the altitude error grows with time. It is hypothesized that the geostrophic wind assumption often becomes a limiting factor depending on the daily weather patterns and in particular on the amount of curvature in the high altitude jet stream. The ability to use meteorological data at a particular locale was not anticipated. Even today, it would be impractical and inaccurate to prepare such data from outside sources. The ability to estimate the radius of curvature as part of an error model in the navigation Kalman filter is a new idea warranting further investigation. Under the assumption of a gradient wind, Blanchard's algorithm is derived along with the partials for a new error state. Test results using actual flight data show the improvement of the model over an ordinary barometric altimeter bias error model.