XMM-Newton Science Analysis System


omsrclistcomb (omsrclistcomb-2.22) [xmmsas_20141104_1833-14.0.0]

Meta Index / Home Page

Source matching

In order to produce the merged source-list from all the individual input source-list files, omsrclistcomb first produces a list of all the input sources sorted into order of decreasing source significance. To decrease the CPU time taken in the matching process, the distance (d) between two sources is computed from their equatorial-rectangulr coordinates (x, y and z), using the following formulae:

• $x=cos(\delta)\times cos(\alpha)$
• $y=cos(\delta)\times sin(\alpha)$
• $z=sin(\delta)$
• $d=\arccos(x_1\times x_2+y_1\times y_2+z_1\times z_2)$

The equatorial coordinates are only computed once.

In order to deal with source-position errors that may be underestimated, and don't in fact take into account ``offsets'' between the coordinates in the different files, each source is assigned a new error based on the following:

• For a source with a likely strong mod-8 pattern (Quality flag bit 4 set), the error in the position is taken to be the maximum of its computed position error and 1 arcsec.
• For extended sources the position error is taken to be the maximum of its computed position error and 0.25 $\times$ its computed semi-major axis, or a minimum of 3 arc secs.

For each source in the list omsrclistcomb then tries to find a counterpart from the other sources in the list, excluding sources from the same source-list file from where the current source originated. Two sources are considered to ``matched'' if the distance between them satisfies the following:

$distance <= \max{(2.5, \min{(4, \textbf{nsigma} * \textbf{error}))}}$ arcsecs

where error is computed from the formula $\sqrt{ err_1^2 + err_2^2}$, where $err_1$ and $err_2$ are the errors in the two source positions and nsigma is the input parameter (default value=3). The matching algorithm may find several possible ``matches'' and will select the closest.