Kalman filtering of wind velocity measurements by use of a first order Gauss-Markov model

A discrete Kalman filter was applied to measure wind velocity using Gauss-Markov (G-M) process to model the random nature of the velocity. The performance of the filter was not strongly dependent on the model parameters. For measurement variance on the order of 1 (m/s)², there was little or no improvement in the estimate of the velocity. For measurement variance below about 0.8 (m/s)², the implementation of the filter degraded the estimate. As the measurement variance increased, the amount of improvement added by the filter approached a constant value.