XMM-NEWTON GO QUESTIONNAIRE

In April 2002, the GOF sent out a questionnaire to US GOs to get general feedback on the XMM-Newton project and specific reasons why so few US lead author papers have appeared in press. We include below the results of the survey, comments from the survey respondents, and our answers in cases where they are appropriate (and where we have any).

In the list below Num is the number of respondents, Ave is the average value of the response, and Dis is the dispersion of the responses. A grade of 0 indicated that the respondent completely disagreed with the statement while a grade of 5 would indicate complete agreement.

Comment	Ave	Dis	Num
I have not (or have just recently) received my data.	3.9	1.4	17
I study extended structures which is currently very difficult.	3.1	1.6	15
he data were too heavily contaminated by soft proton flares to be useful.	2.1	1.4	14
The archive was too sparsely populated.	2.5	1.7	12
The software was difficult to use or too limited.	3.3	1.4	16
Not another %#$*& software system to learn.	3.1	1.3	16
he documentation was difficult to use or too limited.	3.3	1.1	15
had to analyze my Chandra data first because the proprietary period was about to expire and have not had time to get to my XMM data.	2.3	1.5	15
I've given the data to my undergrad/grad student--- They are still working on it (or so they told me).	1.8	1.1	12
I was otherwise occupied/too lazy to analyze the data before now.	1.8	0.8	12
What are you talking about, I have published my data or presented them at a conference.	1.9	1.5	13

Below are a sample of remarks and suggestions which came back with the survey combined with others that have come our way. We will them use them as a forum to address perceptions of certain aspects of the XMM-Newton project.

REMARKS:

• AO-1 data have been slow in reaching GOs, both because of processing delays and because of delays in scheduling of the observations.

This has certainly been unfortunate. AO-1 has extended longer than it was supposed to (although even the original plan had observations extending into 2002) and a number of observations had processing problems. The good news is that all feasible AO-1 observations should be completed by ~February 2003 and the average time between observation and delivery of data set is now less than one month.

• The public archive was delayed in opening and slow in being populated.

This was also unfortunate, and again the situation has improved dramatically. There are now 438 publicly available observation data sets in the XSA.

• AO-2 has been delayed far too long.

Again this was unfor tunate, but the situation is improving. The AO-2 results will be ready for release at the end of July 2002. AO-2 observations will start 2003 February/March and be finished by the end of 2003. With AO-3, the timing of the AO cycle should be nominal, with a cycle length similar to that of Chandra.

• XMM analysis software and documentation require improvement.

While the early versions of SAS were certainly problematic, the more recent releases are functional and reasonably easy to use. The documentation, including the ABC Guide, are available and continue to be improved.

• Many instrument modes are insufficiently calibrated.

The instruments and nearly all modes are typically well calibrated and usable by GOs. While there are some systematics which are still being addressed, e.g., EPIC residuals around the gold edge and silicon line, the calibrations are reasonable for this stage in a mission. The soon-to-be-released EPIC Calibration Guide will aid users in interpreting their data.

• The proposal submission process needs improvement.

For AO-3 the proposal submission process will be changed to have two stages. Only successful proposers will need to enter the wealth of instrument configuration data necessary for an observation. The GOF is working with the SOC to provide a version of RPS for the Stage 1 proposal submission. This should remove most of the pain and anguish from the process.

• European data sets appear to be processed more promptly than US data sets.

This is not true.

• The ABC Guide has been helpful.

Thanks. Suggestions for improvements are appreciated.

• SAS is RAM intensive and slow. This will be a problem for users without excellent computing resources.

$4,000 will now buy a dual 1 GHz processor Linux box with 2 Gb RAM. With such a machine and the improved SAS (which is significantly faster), processing pains for the most part should go away.

SUGGESTIONS:

• Separate the Chandra and XMM-Newton proposal processes in time.

An issue for the XMM-Newton Users' Group Commitee (USG).

• Coordinate the Chandra and XMM-Newton missions to maximize their scientific return.

An issue for the USG and the CXC.

• A policy is needed for the reobservation of targets where the original data are too heavily contaminated for the proposed science.

An issue for the USG.

• XMM-Newton GOs should be funded at the same rate as Chandra GOs.

With our success in the Senior Review we should be able to do so. We are extremely pleased by this result.

• Individual scientists at the GOF should install SAS software, analyze XMM data, and make sure that the on-line information is correct and up-to-date, based on their own experiences.

This is what we do.

• More cookbooks and fully worked examples would be helpful.

We are looking into this and will try to provide such.