A chaotic attractor in Timing noise from the vela pulsar?

We have analyzed 14 years of timing residual data from the Vela pulsar in order to determine if a chaotic dynamical process is the origin of timing noise. Using the correlation sum technique, we obtain a dimension of ∼ 1.5. This low dimension indicates underlying structure in the phase residuals which may be evidence for a chaotic attractor. It is therefore possible that nonlinear dynamics intrinsic to the spin-down may be the cause of the timing noise in the Vela pulsar. However, we have found that the stimulated random walks in frequency and frequency derivative often used to model pulsar timing noise also have low fractal dimension, using the same analysis technique. Recent work suggesting that random processes with steep power spectra can mimic strange attractors seems to be confirmed in the case of these random walks. It appears that the correlation sum estimator for dimension is unable to distinguish between chaotic and random processes.