Short-window spectral analysis using amvar and multitaper methods

In this paper, we compare two popular methods for estimating power spectrum for short time series, namely adaptive multivariate autoregressive (AMVAR) method and the multitaper method. By analyzing a simulated signal (embedded in a background Ornstein-Uhlenbeck noise process) we demonstrate that the AMVAR method performs better for very short data when compared to the multitaper method. We also show that coherence can still be detected in noisy bivariate time series data even if the individual power spectra fail to show any peaks.