Search for Millisecond Pulsars for the Pulsar Timing Array project

Pulsars are rapidly rotating highly magnetised neutron stars (i.e. ultra dense stars, where about one solar mass is concentrated in a sphere with a radius of ~ 10 km), which irradiate radio beams in a fashion similar to a lighthouse. As a consequence, whenever the beams cut our line of sight we perceive a radio pulses, one (or two) per pulsar rotation, with a frequency up to hundred of times a second. Owing to their compact nature, rapid spin and high inertia, pulsars are in general fairly stable rotators, hence the Times of Arrival (TOAs) of the pulses at a radio telescope can be used as the ticks of a clock. This holds true in particular for the sub­class of the millisecond pulsars (MSPs), having a spin period smaller than the conventional limit of 30 ms, whose very rapid rotation and relatively older age provide better rotational stability than the ordinary pulsars. Indeed, some MSPs rotate so regularly that they can rival the best atomic clocks on Earth over timespan of few months or years.This feature allows us to use MSPs as tools in a cosmic laboratory, by exploiting a procedure called timing, which consists in the repeated and regular measurement of the TOAs from a pulsar and then in the search for trends in the series of the TOAs over various timespans, from fraction of seconds to decades.For example the study of pulsars in binary systems has already provided the most stringent tests to date of General Relativity in strong gravitational fields and has unambiguously showed the occurrence of the emission of gravitational waves from a binary system comprising two massive bodies in a close orbit. In last decades a new exciting perspective has been opened, i.e. to use pulsars also for a direct detection of the so far elusive gravitational waves and thereby applying the pulsar timing for cosmological studies. In fact, the gravitational waves (GWs) going across our Galaxy pass over all the Galactic pulsars and the Earth, perturbing the space­time at the pulsar and Earth locations, as well as anywhere along the line­of­sight from the Earth and each of the pulsars. This in turn produces a modulation in the rhythm of the TOAs of the pulses from all the pulsars, with the variation in the TOAs having a strength which is proportional to the amplitude of the GW and a periodicity related to the frequency of the GW. Of course if they are caused by a common physical phenomenon (like a passing­by GW), these variations of the TOAs are expected to be somehow correlated between the various pulsars, allowing us to disentangle this effect from other effects which could mimic the occurrence of such modulation, like intrinsic irregularities in the rotation of a pulsar, changing interstellar medium along the line of sight, error in the reference clocks used for determining the TOAs and so on.The consideration of the aforementioned possible sources of additional effects which could mask the signature of a genuine GW shows that a safe direct detection of a GW cannot involve the observation and timing of a single pulsar. Instead, it has been theoretically shown that high precision timing over a 5­10 years data­span of a network of suitable MSPs forming a so­ called Pulsar Timing Array (PTA) ­ in which the pulsars are used as the endpoints of arms of a huge cosmic GW detector ­ would allow us to overcome the previous problems and open the possibility of a direct detection of GWs. In particular such apparatus is able to detect GWs in the frequency range between 10 ­9 and 10­7 Hz, with the best sensitivity around the nanoHz. Given the frequency range of operation, the most favorable source of GWs for a PTA appears to be the cosmological background of GWs produced by the coalescence of supermassive binary black­holes in the early stages of the Universe evolution, at redshift around 1­2. In order to set up a suitable PTA it is necessary on one hand to search for new MSPs having the required clock stability and signal intensity, and on another hand to perform regular high­precision timing observations of the available sample, combining the results from all the pulsars with the use of a solid and well tested software, capable of revealing the genuine GW signal which is searched for. This work focuses on the first task, in an attempt to enlarge the number of suitable MSPs, in the framework of the High Time Resolution Universe (HTRU) survey for pulsars and fast radio transients, that is currently underway at the 64­m Parkes Radio Telescope (NSW, Australia). This experiment has been designed in 2007 and started three years ago, with the main scope of largely increasing (possibly doubling) the total number of MSPs known in the Galactic Field (there were only about 40 of them until 2009). The enlarged sample may provide some very good MSP­clocks to be added to the still relatively poor list of objects well suited for belonging to a PTA. In the first chapter of this thesis an overview of the pulsar phenomenon is given, with also a description of the timing technique and its physical applications. The search methods that can be used to analyse the data in order to find isolated and binary pulsars are reported in the second chapter. The third chapter describes part of the work performed by me in the framework of the HTRU survey; in particular the search for MSPs in the HTRU data with a data reduction pipeline sensitive also to highly relativistic systems (i.e. to binary pulsars in close orbits). While performing the aforementioned search, it emerged the issue of the inspection of the hundreds of thousands of pulsar candidates produced by the adopted pipeline, the vast majority of them being the result of radio interferences. Therefore, a new approach has been explored for making manageable the human intervention in the procedure of selection of the trustable candidates, namely the use of an Artificial Neural Network on the pulsar candidates. The fourth chapter is devoted to report on that. At the end, a brief summary of this thesis work is given, as well as a list of the publications, in preparation and resulting from the HTRU collaborative effort.