|
Figure 7 shows another part of the prepfold plot, and this one is much different than the previous two. The x-axis is labeled as "DM", and the y-axis as "Reduced $\mathbf{\chi^2}$}". First, let's talk about the DM. DM stands for Dispersion Measure, and it is a very important characteristic of our pulsar signal. The dispersion measure doesn't actually have anything to do with the pulsar itself. Instead, the dispersion measure tells us something about the space between Earth and the pulsar. We often think of space as being empty, but that isn't quite true. Space is filled with many things, though the density is so low that compared to Earth, space is almost empty. One of the things that we find in space are electrons &emdash subatomic particles that normally orbit the nucleus of an atom. These electrons disperse the pulsar's signal (hence the name "dispersion measure"), causing lower observing frequencies to arrive later than higher observing frequencies. The electrons can also scatter the signal much the same way smoke scatters visible light. The dispersion measure is a way of telling us how many electrons the signal encountered on it's way to Earth. The larger the dispersion measure, the more electrons the signal encountered. This could happen for two reasons &mdasdh either the pulsar is very far away, or the density of electrons in the space between Earth and the pulsar is relatively high. Both will cause an increase in the dispersion measure.
What about the "Reduced $\chi^2$"? First off the symbol "$\chi$" (pronounced "Kigh", or /kaI/ in the phonetic alphabet) is simply a Greek letter "X". The Reduced $\chi^2$ is a statistical value that indicates the goodness of a given measurement. If you are familiar with statistics, then you know that the Reduced $\chi^2$ indicates how well a given model matches some actual data. A large value of Reduced $\chi^2$ means that the model and data are not in good agreement, and a Reduced $\chi^2$ of one means that they are. We use the Reduced $\chi^2$ in an uncommon way. Our model doesn't include a pulsar, because we don't know the properties of the pulsar ahead of time. So when we compare this pulsar-less model to data that does contain a pulsar, we see a high Reduced $\chi^2$, which is exactly what we want.
Now that we know what are axes are, what does the dispersion measure plot tell us? The peak of the curve tells us the most likely value of DM. Ideally, we would like to see a sharp peak at one value of DM. This would mean that our DM is pretty well measured. If the peak is broader, it means that our DM measurement is less certain.
It is important to note where the DM curve is peaking when you are trying to determine if your pulsar candidate is real. If the curve reaches its highest point at a DM of zero, then the signal must not have traveled through any electrons to get to E arth. But that must mean that the signal actually originated at the Earth, and since there are no pulsars on the surface of our planet (a very good thing!) then we know that we must actually be looking at RFI.