Alexandra (Wheaton Academy)
Mentor: Chris Stoughton
This summer I interned with Chris Stoughton and assisted him with running diagnostics on the microwave kinetic inductance detector. The MKID is a new photon detector that is able to record the energy of incoming photons, as well as the time in which they were detected. In a normal data stream, one or two photons are detected per microsecond. However, there are intervals of time in which thirty, seventy, or more photons are detected. These events are dubbed 'cosmic ray events'. Such events may be caused by actual cosmic rays, dark matter, or interference from the magnet that is part of the detector. My task was to use the programming language Python to mask out cosmic ray events so that the best data possible is available for the creation of spectra and images of celestial objects. My program makes a histogram of the photons per microsecond, then fit this histogram with a gaussian function. Then, it calculates the mean of the gaussian and the standard deviation. The interval that gets masked out is the mean minus the sigma multiplied by a constant to the left, and double that amount to the right. More data is cut out to the right because there tends to be more bad data to the right of the gaussian's mean. After the cosmic events are removed, the data fits a very nice poisson distribution. The mean of this distribution is about 1.5, which indicates that it is most likely that one or two photons will be detected per microsecond. This is consistent with a normal data stream for the detector, which means that my masking of cosmic rays was effective.