Events & Event lists

Events are any occasion in which there is a signal remaining in a CCD pixel above the event recognition threshold after bias subtraction. Thus an event may be produced by X-ray photons, charged particles, noisy electronics, stray optical light or defective CCD pixels, just to name a few. It is part of the scientist's analysis task to determine how to maximize the utility of the data by appropriately selecting events so as to maximize the signal and minimize the background.

Before events are extracted, each raw FITS image is bias and overclock corrected. This is done in three steps.

1.

First the average bias image is subtracted from every pixel in the raw FITS image (image, ${\cal I}$, and overclock, ${\cal O}$, pixels).

$$\acute{p}(x,y,n)=p(x,y,n) - \bar{p}(x,y) \ \ \ \forall (x,y) \in \{ {\cal I, O} \}$$ (3.7)

2.

Then the overclock pixels associated with each quadrant, q, are averaged

$$o(q,n)=\frac{1}{n(q)}\sum_{(x,y)\in {\cal O}(q)} \acute{p}(x,y,n) \ \ \ \forall q=\{ {\rm A,B,C,D} \}$$ (3.8)

where n(q) is the number of overclock pixels in each quadrant

3.

And finally the overclock correction is added back

$$\grave{p}(q,n)=\acute{p}(x,y,n) + o(q,n) \ \ \ \forall (x,y)\in {\cal I}(q) : q=\{ {\rm A,B,C,D} \}$$ (3.9)

These bias corrected pixel values are then used for event detection and subsequent extraction. Note that the reason for the overclock pixels is to allow removal of short term drift in the electronic baseline on the timescale of CCD row readout.

Depending on the flux of the source, events can be extracted in one of two ways: photon counting mode and integration mode. Both of these are described in detail below.