Dear Nick, Gigi, COSMOS cluster folks, 
Cc: COSMOS fellows

Last week we had a cluster workshop and a lot of couloir discussions wrt how to
proceed further with cluster finding, list sharing, follow-up and so on. I think
we miss here a well-defined strategy, which could make one of the topics of an
upcoming telecon discussion. Here I would like to give an outline of the methods
we plan to apply to the COSMOS data and the timescale for getting the results.

(1) Photo-z galaxy concentration vs any X-ray source. 
Status: performed for the existing X-ray fields. Need to sort out a scheme of
cluster follow-up as well as data sharing. Before I do something, m.b. we should
discuss how to orgainize a joint work on that, so people do not feel affended.
Guenther Hasinger suggests that the proposal should stem from SUBARU people. One
settled, should take a week to release a list.
Advantages: should include also clusters strongly affected by AGNs.
Disadvantages: may include filaments and chance allignment 


(2) Wavelet-scale search for the X-ray source extent (Vikhlinin et al. 1998)
Status: takes me 1 month to implement given the current work load
Advantages: distant clusters that are not in photo-z catalog, less confusion
Disadvantages: need to combine with more sophisticated cluster list cleaning
algorithms.

(3) Grouth-curve analysis
(a) individual sources
Status: Takes 1/2 year to implemet given the current work load for me and Hans
Boehringer
Advantages: 1d - can use stability of the 1d PSF profile of XMM and its nearly
position independence (e.g. directly applicable to mosaics)
Disadvantages: cannot distinguish multiple sources

(b) multiple sources
Status: takes 1 year to fully implement
Advantages: the most sophisticated test, known to yield the cleanest cluster
lists possible
Disadvantages: Needs 2d PSF, which is observation-specific (tangential
elongation wrt telescope center, telescope specific) and therefore requires
XMMSAS programming on one hand and understanding the quality of the 2d PSF
calibration on the other.


(4) X-ray colors
Status: color images from Guenther Hasinger; I propose to combine it with the
method #1; available on the timescale of 1 month
Advantages: additional support using the nature of source emission
Disadvantages: should be used in combination with other methods.

(5) combined search for clusters in optical and X-ray data
Status: implemented by Schuecker et al. (2004); no plans to apply to COSMOS
Advantages: lower limits achieved
Disadvantages: XMM is background limited, not count limited. Therefore validity
of the X-ray source needs to be proven before further complex analysis could be
undertaken.

Further suggestions are most wellcome.


Cheers,
	Alexis.