# Posts Tagged spectral leakage

## Recent Postings from spectral leakage

### Optimal strategies for continuous gravitational wave detection in pulsar timing arrays

Supermassive black hole binaries (SMBHBs) are expected to emit continuous gravitational waves in the pulsar timing array (PTA) frequency band ($10^{-9}$–$10^{-7}$ Hz). The development of data analysis techniques aimed at efficient detection and characterization of these signals is critical to the gravitational wave detection effort. In this paper we leverage methods developed for LIGO continuous wave gravitational searches, and explore the use of the $\mathcal{F}$-statistic for such searches in pulsar timing data. Babak & Sesana 2012 have already used this approach in the context of PTAs to show that one can resolve multiple SMBHB sources in the sky. Our work improves on several aspects of prior continuous wave search methods developed for PTA data analysis. The algorithm is implemented fully in the time domain, which naturally deals with the irregular sampling typical of PTA data and avoids spectral leakage problems associated with frequency domain methods. We take into account the fitting of the timing model, and have generalized our approach to deal with both correlated and uncorrelated colored noise sources. We also develop an incoherent detection statistic that maximizes over all pulsar dependent contributions to the likelihood. To test the effectiveness and sensitivity of our detection statistics, we perform a number of monte-carlo simulations. We produce sensitivity curves for PTAs of various configurations, and outline an implementation of a fully functional data analysis pipeline. Finally, we present a derivation of the likelihood maximized over the gravitational wave phases at the pulsar locations, which results in a vast reduction of the search parameter space.

### The impact of a stochastic gravitational-wave background on pulsar timing parameters

Gravitational waves are predicted by Einstein’s theory of general relativity as well as other theories of gravity. The rotational stability of the fastest pulsars means that timing of an array of these objects can be used to detect and investigate gravitational waves. Simultaneously, however, pulsar timing is used to estimate spin period, period derivative, astrometric, and binary parameters. Here we calculate the effects that a stochastic background of gravitational waves has on pulsar timing parameters through the use of simulations and data from the millisecond pulsars PSR J0437–4715 and PSR J1713+0747. We show that the reported timing uncertainties become underestimated with increasing background amplitude by up to a factor of $\sim10$ for a stochastic gravitational-wave background amplitude of $A=5\times 10^{-15}$, where $A$ is the amplitude of the characteristic strain spectrum at one-year gravitational wave periods. We find evidence for prominent low-frequency spectral leakage in simulated data sets including a stochastic gravitational-wave background. We use these simulations along with independent Very Long Baseline Interferometry (VLBI) measurements of parallax to set a 2–sigma upper limit of $A\le9.1\times 10^{-14}$. We find that different supermassive black hole assembly scenarios do not have a significant effect on the calculated upper limits. We also test the effects that ultralow–frequency (10$^{-12}$–10$^{-9}$ Hz) gravitational waves have on binary pulsar parameter measurements and find that the corruption of these parameters is less than those due to $10^{-9}$–$10^{-7}$ Hz gravitational waves.

### On detection of the stochastic gravitational-wave background using the Parkes pulsar timing array

We search for the signature of an isotropic stochastic gravitational-wave background in pulsar timing observations using a frequency-domain correlation technique. These observations, which span roughly 12 yr, were obtained with the 64-m Parkes radio telescope augmented by public domain observations from the Arecibo Observatory. A wide range of signal processing issues unique to pulsar timing and not previously presented in the literature are discussed. These include the effects of quadratic removal, irregular sampling, and variable errors which exacerbate the spectral leakage inherent in estimating the steep red spectrum of the gravitational-wave background. These observations are found to be consistent with the null hypothesis, that no gravitational-wave background is present, with 76 percent confidence. We show that the detection statistic is dominated by the contributions of only a few pulsars because of the inhomogeneity of this data set. The issues of detecting the signature of a gravitational-wave background with future observations are discussed.