The Parkes Radio Telescope (photo courtesy Andrew Cameron)

The HTRU-S Lowlat project comprises one part of the “High Time Resolution Universe” HTRU survey, an all-sky, high time resolution blind survey focusing on the discovery of new radio pulsars as well as other radio transient objects. HTRU-South (HTRU-S) covers the southern regions of the sky, with observations taken at the 64m Parkes Radio Telescope in Australia, while a corresponding survey, HTRU-North (HTRU-N) covers the northern regions of the sky with observations taken using the 100m Effelsberg telescope in Germany. Together these two surveys, combined with advances in instrumentation and digital signal processing, provide a complete picture of the sky at unprecedented time resolution and sensitivity.

The HTRU-S survey strategy has been to break down down the sky into three different survey regions with respect to Galactic latitude: high- latitude, mid-latitude and low-latitude, with the observations in each region tailored to achieve specific scientific goals. The low-latitude survey presented here comprises a region of the sky centred on the Galactic plane (|b| < 3.5
°). As this region of the sky contains some of the densest regions of the interstellar medium (which could potentially obscure or weaken the signal from any interesting objects), it contains the longest observations of the HTRU-S survey, with each one lasting 72 minutes. This extra boost in sensitivity will probe the Galaxy deeper than every before and aid in uncovering pulsars and other objects too faint to have previously been detected by other surveys.

Science and Goals

The HTRU project's primary scientific goal is to allow for the discovery and investigation of new and scientifically-interesting pulsars, along with other radio-transient objects. The unique properties of pulsars make them useful as scientific laboratories. Their intense gravitational forces and magnetic fields make them highly useful in the study of fundamental physics. The highly-precise timing of the pulses from many of the so-called Millisecond Pulsars (MSPs) has lead to the formation of large pulsar timing networks, aiming to uncover the first signs of Gravitational Waves. The discovery of new pulsars will provide us with more opportunities to explore these and other fundamental areas of science.

In particular, the low-latitude Galactic plane region surveyed by HTRU-S Lowlat is where the greatest number of relativistic pulsar binary systems are expected to be found. Many pulsars are found in orbit around other celestial objects, and those orbiting in tight binary systems with other neutron stars (or potentially, a black hole) are of very high interest, as the intense gravitational interactions allow for some of the most stringent tests of General Relativity, as well as other gravitational theories. The best example of such a system is the so-called Double Pulsar, PSR J0737-3039A&B. One of HTRU-S Lowlat's primary goals is the discovery of more such relativistic binaries.

  • What is a pulsar?

First discovered in 1967, pulsars are rapidly-rotating, highly-magnetised neutron stars, the collapsed cores of earlier stars that long ago exploded into supernovae. They have masses on the order of our own Sun, and yet in astronomical terms they are tiny, with diameters of only 10 - 20km. The primary observational feature of a pulsar are the jets of radiation originating from its magnetic axis. As the pulsar rotates (with rotational periods currently ranging anywhere from ~1.4ms to 11.7s), these jets sweep through space like a cosmic lighthouse, and can be detected on Earth as a series of regular radio pulses. To date, at least 2400 pulsars have been discovered.


All observations for the HTRU-S survey have been carried out using the Parkes 21-cm Multibeam Receiver, which is capable of observing 13 beams simultaneously. This both increases survey efficiency and allows for better cross-correlation and removal of radio-frequency interference (RFI). This receiving front-end is combined with a digital processing back-end, the Berkely-Parkes-Swinburne Recorder (BPSR), which performs real-time analog-to-digital conversion of the incoming telescope data and is what allows for the HTRU-S survey's unprecedented time resolution (64 μs) across 1024 frequency channels.

Data Processing

The sheer volume of data produced by the HTRU-S Lowlat project presents a serious computational challenge. The over 1200 observations taken as part of the project make up around 263 TB of data, all of which has to be thoroughly processed. To this end, we have made use of several supercomputing facilities. Presently, our primary facilities is the 4488-core cluster 'Hercules' operated by the Max Planck Computing & Data Facility in Garching, Germany. Previous processing has been carried out at the  Australian National Computational Infrastructure (NCI) located at The Australian National University (ANU) on the 57,472-core high-performance distributed-memory cluster 'Raijin', as well as on the now decommissioned 1492-node Sun Constellation cluster 'Vayu' and the 156-node SGI cluster 'XE'. We have also used the 1456-core supercomputer 'HYDRA' located at the Jodrell Bank Observatory and the 64-core computer 'Miraculix' at the Max-Planck-Institut für Radioastronomie.

Follow-up Observations

Following the detection of a candidate pulsar amongst the survey data, follow-up observations are first conducted with the Parkes Radio Telescope in order to confirm the pulsar's existence and position. Following this confirmation, the pulsar is added to a roster of timing observations which are carried out either with the Parkes Radio Telescope or, for those pulsars which are sufficiently north in the sky so as to be visible, with the 76m Lovell Telescope at the Jodrell Bank Observatory in the UK. These timing observations allow for the properties of the pulsar to be determined with high precision, allowing for future science to be conducted using the pulsar.

Technical Specifications

Parameter Value
-80° < l < 30°
|b| < 3.5°
Area of sky surveyed (in Galactic coordinates)
τobs (s)
4300 Length of each observation
Nbeams 13
No. of telescope beams per observation
τsamp (μs) 64
Time resolution / Sampling time
νcentral (MHz)
Central frequency
B (MHz) 340*
Total bandwidth
Δνchan (kHz) 390.625
Channel bandwidth
Nchans 870*
No. of channels
Data length (samples)
No. of samples per beam per observation
Data / beam (GB)
~16 Size of data per beam per observation

* Recorded bandwidth is at 400 MHz with 1024 channels, centred on 1382 MHz, but the usable band is reduced due to satellite-induced RFI.