Muon Signals in Igloos

I have tried to find some characteristics in the raw Milagrito data to distinguish the signals of muons in the igloos from background. In essence I have been making histograms of igloo data (adc, low and high TOT) with various cuts to try to resolve and enhance any features that could be interpreted as muons.

So far I have tried three main types of cuts, in various combinations, to get these signals. The three cuts are shower angle, core location, and what we have affectionatley termed the "anti-Scott" or ("Sbar") cuts. This cut requires that an igloo's signal for a given event be read out only if it's neighbors are empty i.e. if it's neighbors have no signal for that event. This is meant to increase the odds that an igloo's signal is due to a muon and not due to punch-through from a very energetic shower. The angle cut (theta < 20 degrees) is meant to restrict the path of through-going muons. This will help ensure that the muons we see are indeed associated with air-showers. And the core location cuts (generally: core > ~20m from igloo) are meant to increase the ratio of muons to other shower particles, since the lateral distribution of muons is flatter than that of electrons.

I have been running my cuts on about 20M events (one DLT) and the results are not discouraging. Take a look at some histograms for igloo 4 and igloo 7, and decide for yourself. In these graphs, the histogram number tells you what the contents are, as follows. The first three digits tell you the daq channel number. Thus any histogram of data from igloo 4 is 244##. The fourth digit designates the type of data in the histogram, 1=adc, 2=low TOT, and 3=high TOT. And the last digit tells which cuts were made. In these figures I show only plots with no cuts (last digit=0), with the Sbar cut (1), and with all three cuts (7).

The following features are, I think, pertinent. In the ADC graphs, for each igloo, there is a slight shoulder about 600-700 bins above pedestal. This shoulder is accentuated by the cuts. Additionally, in the high TOT graphs, there is a peak near bin 400, which is also accentuated by the cuts. Isabel tells me that 350-400 TOT counts is consistent with 30 or so PEs, so this seems reasonable given the QVT results.

It is possible to fit Gaussian curves to these features, both ADC and TOT. Examples are shown at right. I fit the "shoulders" in the ADC plots to the sum of an exponential and a Gaussian. Typically, I get a free calibration method for the ADC plots because they contain pedestal and one PE peaks. I use the separation between these two peaks to give me a conversion rate for ADC bins to PEs, and then use this to calculate the number of PEs in each shoulder region. I do this for each igloo and each cut. The results are consistent with previous QVT studies, and they are self consistent.