Characterizing EDGES Radio Frequency Interference

Hello LoCo followers! As the avid readers should know, the LoCo team is working on developing a ultra-sensitive radio telescope, EDGES, in order to detect faint emission signals from the very earliest stars and black holes that formed in the universe. Since the signals are so faint, EDGES must be the most sensitive instrument of its kind. However, the high response of the telescope to small signals can be disadvantageous, since there are many other radio sources that can “drown out” the desired signal. Most everyone has experienced this effect as static on a car radio. In this case, the EDGES frequency range actually includes the FM band, and the radio signals that you want to hear in your car is, to us, just like ‘static’ that interrupts your favorite broadcasts! In fact, astronomers call this unwanted interference ‘noise’ even though it is caused by electromagnetic waves instead of sound waves.
Here at ASU, I am looking at the radio frequency interference (RFI) patterns that disturb our observations of the early universe. In figure 1, you can see a typical plot of the intensity at a given frequency during the course of the day. The X-axis is a plot of frequency, and the Y axis shows time at one minute intervals. A red pixel shows that there was a lot of incoming flux at a given frequency, and blue shows a low level of flux. The large red sphere is actually the radio emission of our Milky Way galaxy. The vertical bands of high flux correspond to radio frequency channels that we humans use for broadcasting. Some of these bands are used for FM radio, some are used for GPS, and others are used for industrial or amateur radio communication (like cell phones or satellite TVs).

Antenna Temp 2011_315_00_Ta

The intensity of the RFI signals in these bands changes throughout the day for a wide variety of reasons. I am looking for times when the power is unusually high, and seeing if those events correlate to known environmental phenomenon. For instance, one of Earth’s atmospheric layers is composed of high temperature particles that have been stripped of their electrons and form ions, thus the layer’s name ‘ionosphere’. When a force perturbs this layer, the charges move around in wavelike patterns. Moving electrical charges radiate photons, and the frequency of the radiation is dependent on the frequency that the ions oscillate. If the ions oscillate at radio frequencies, then there will be an increase in radio photons coming from the ionosphere, and the EDGES instrument will record that event. Other RFI sources that cause noise in the data are meteor showers (that perturb the ionosphere), ionospheric clouds, solar flares, and geomagnetic storms.
The frequency at which radio waves can propagate through the ionosphere is called the critical frequency. Signals at higher frequencies than this escape to space, while signals at lower frequencies
are reflected back towards Earth. Figure 2 shows a plot of the average critical frequency of the f1 layer of the ionosphere for each day from 2007 to present, which spans the time period that the EDGES instrument has been running. You can see that there are seasonal peaks each year, due to the orientation of the earth’s magnetic field compared to the direction of the sun. I am interested in finding the spikes in the data, to see if there are any peaks in the RFI intensity on those dates.

RFI 2011_315_00_rfi

As mentioned above, infalling meteors are known to cause perturbations in the ionosphere. Many of these meteor showers are spectacular to view with the naked eye, but a few are so faint that we only know they are there because of the disturbances they make in the ionosphere that have been recorded by other radio instruments. Figure 3 shows a plot of how many meteors are expected to hit the atmosphere for any given day. They are separated into ‘visible’ and ‘radio’ showers, based on whether they are visible to the naked eye or have been detected by radio instruments only. The X axis runs over one year only, since the Earth only runs into each group of meteors once per year due to our orbit around the sun. The Y axis gives the Zenith Hour Mean (ZHM), which is the total number of events predicted to occur during the hour that the shower has the most meteors.

Total Expected Radio and Visible

So far, I have been working on getting the data into a format that will be meaningful to do the comparison between. In the next few weeks I should start to see if there are any of the correlations between the ‘noise’ in the EDGES data and the events that happen in the sky above it. If this is successful, I can use the information to remove the events from the data so the astronomical signal we want to detect will be clearer. Wish me luck!