The Chandler wobble (
cw) is the most important rotational eigenmode of the Earth. Its parameters, the frequency F (or, equivalently, the period T=1/F) and the quality factor Q, are important for studying global dynamics of the Earth because they define the equation describing geophysical excitation of polar motion and are closely related to various geophysical parameters. Our purpose here is to derive the improved estimate of the
cw parameters by using the stochastic models of the polar motion and geophysical excitation data. We apply the following three algorithms of data analysis: 1) Autoregressive (
ar) model of the observed polar motion fitted by imposing the maximum entropy condition; 2) Autoregressive integrated moving-average (
arima) model which takes into account the influence of the polar motion observation errors, estimated by applying the maximum likelihood condition; 3) Kalman filter developed on the base of stochastic models of both the polar motion observations and the excitation data. This filter can be applied to analyze either the polar motion data alone or simultaneously the polar motion and geophysical excitation data. These three algorithms are used in this work to derive the
cw parameters from analysis of different available sets of polar motion and geophysical excitation data. The estimated values of F and Q are compared to each other as well as to the results known from the literature.