As you may have begun to realize, there is an awful lot of work done on computers when it comes to finding pulsars. There are many different packages of software designed for this task, and the one you are going to use is called PRESTO. PRESTO takes the Fourier transform of a data set and uses some clever algorithms to search for pulsars. When PRESTO finds a promising candidate, it can go back to the time domain data and use a program called prepfold to fold the data at the period of our candidate. What comes out is a plot like the one in Figure 4. There is a lot of information on this plot, and at first it can look pretty intimidating. So let's take each piece of the plot and break it down, explaining what it shows us and why the information is important. When we put it all together, you will be able to look at one of these plots and decide if it shows a true pulsar.
|
|
|
Figure 5 shows one of the most important parts of a prepfold plot. This is the time domain and pulse profile. The time domain plot shows us how the strength of our signal varies during the course of our observation. Each gray square represents the strength of the signal in some small piece of the data (we refer to this small piece as a bin). The darker the bin, the stronger the signal. A white bin indicates that no signal was detected in that bin. This plot is a representation of our folding process. Each row on this plot is a sub-fold, i.e. a fold of a small chunk of our data this is a few seconds long. This small chunk is simply the "height" of a bin, and is called a sub-integration. The "width" of each row is equal to the length of the fold, which is nothing more than the period of our candidate. If you go back to Figure 3, you will recognize a sub-fold as being a few layers of our folded paper.
The y-axis is labeled as "Time (s)" and is simply the time in seconds since the start of our observation. The x-axis is labeled as "Phase" and goes from 0 to 2.0. The phase tells us how far we are through the pulsar's rotation. A phase of 0.5 means that the pulsar has gone through half a rotation; a phase of 1.0 indicates one full rotation; a phase of 1.5 indicates one and a half rotations, and so on. Each sub-fold is equal to one full rotation, so showing the phase going from 0.0 to 1.0 is the same as showing a full sub-fold. We have shown two full rotations for clarity--if we only showed one rotation, an the pulse landed just at the end of this rotation (i.e. at a phase of 1.0) then it would be on the edge of our plot and hard to see.
Every time the beam passes in front of us, we will see an increase in the strength of our signal. Furthermore, the beam will always pass in front of us at the same phase. This explains why there is a dark line in the time series. It is telling us that, from the beginning of our observation all the way until the end, we got a sudden increase in the strength of our signal, and that this always occurred at the same phase.
Now let's look at the top of this plot. This is known as the pulse profile. and it gives us the strength of the pulse as a function of phase. This is nothing more than the result of our full folding process. To make it, we add up, or integrate, the results of all the sub-folds. The peak corresponds to when the pulsar is "on", and the random lines that you see when we are "off-pulse" is the background noise. Ideally, we should see a pulse that is much stronger than the noise level, indicating a strong pulsar. If the pulse is not much stronger than the noise, then we simply can't say with confidence that we have a real detection.