Chandra Users' Committee Recommendations to Chandra Director Chandra Users' Committee Meeting, June 12-13, 2000 1) Monthly Telecon: The Users Committee was concerned that the interval between meetings is too long considering the rate at which important issues arise. We feel that progress on issues of concern to users might be more rapid if we could monitor developments on a shorter time scale. Thus, we would like to have a monthly one-hour long telecon with the director and key members of the CXC. We would provide a short agenda of key concerns in advance. The Director could also use these telecons to keep the CUC informed on important developments with Chandra. 2) Extended Source Spectra with ACIS: There is a desperate need for software to deal with ACIS spectra from extended sources. The CXC needs to provide software for extracting spectra and calculating response matrices, either by developing it in-house or by actively facilitating the distribution of software developed by users. This needs to be done immediately. The Users Committee wishes to draw attention to the fact that at the current rate of software development, the analysis of extended source spectra may be impossible prior to the public release of Cycle-1 GO data. This will most certainly cause very high levels of unhappiness in the community. 3) Grating Analysis: Grating spectroscopy is one of the great strengths of CXO. At present, users have not been provided with a software system which covers the basic grating analysis requirements. Grating tools which were released as part of CIAO version 1 contained numerous bugs, some of which caused silent but fatal errors to be introduced into user-extracted data products. Some obvious and basic tools are missing altogether. We fear that the current rate of progress on basic grating tools and data analysis threads is so slow that much of the grating data will remain unanalyzed for some time. Again, we are concerned that GO data will go public before the PIs have any chance to analyze their data. This can only result in high levels of unhappiness in the community. The level of effort in this area (at CXC) seems to be small, even though this specific concern was raised in previous CUC meetings. The current style is to develop all software under the CIAO umbrella. While this is perhaps a desirable approach in the long term, the current need is sufficiently urgent that one may wish to consider a more flexible approach, e.g. using tools developed by CXC people at MIT. With grating data, extended source spectra with ACIS, timing data and other issues, CXC faces a danger than unless it can provide necessary software expeditiously, users may conclude that CIAO is not useful, and abandon it. 4) Keeping the Community Informed: The CUC feels that it is imperative that the CXC keep the community informed of important developments with Chandra. We suggest that the CXC provide this information through a standard E-mail exploder, as well as through the CXC webpage. This is the standard practice with other observatories (e.g., ROSAT and HST). An example of such information might be informing the community as soon as possible about the HRC timing problem, now that the Cycle-2 proposal deadline is past. 5) Director's Discretionary Time: An up-to-date list of targets approved under this program should be displayed more prominently on the Chandra webpages. Approved DDT programs should also be publicized via the E-mail exploder discussed above. The point is that DDT observations have either no or only a very short (3 month) proprietary period, are likely to be of great interest to the X-ray astronomy community, and don't appear on the standard GTO/GO target lists. At every CUC meeting, the Director should present a summary of all DDT requests and the subsequent dispositions (accept/reject). 6) HRC Timing Problem: The CUC felt that the handling of information on the HRC timing problem during the Cycle-2 proposal period was reasonable. However, in the future we suggest that the Chair of the CUC be consulted about similar problems. The X-ray community should be informed of this problem as soon as possible via the CXC webpage and the E-mail exploder. This is to make sure that no one (either a PI of an HRC observation or another user analyzing archive calibration data) is unaware of this problem. For the near future, it is critical that the HRC team proceed rapidly to carry out the necessary tests of remedial techniques to recover the msec timing capability of the HRC. The Chair of the CUC should be kept informed of the progress of these tests. On the technical side, we are concerned that the proposed tests are not rigorous enough to evaluate the HRC at the 3-4 millisecond level. We would like observations of one or more bright, faster pulsars to determine timing precision at the msec level. These might involve GO/GTO observations which the CXC would check for timing accuracy but would otherwise remain proprietary. Regardless of the above tests and demonstrations, the CXC should immediately carry out technical reviews of all HRC Cycle-2 proposals that involve timing in order to determine whether they are feasible with the HRC in its present condition, feasible using the remedies proposed, feasible using alternatives (ACIS) or simply not feasible. This information needs to be provided to the review panels prior to the review. The CXC should contact all HRC PIs (with completed Cycle-1 observations) and enquire whether their primary science goals (as identified in their proposals) were critically compromised by the msec timing problems of the HRC. Assuming that appropriate solution[s] have been identified, these PIs should have the option of sending a justification (1 page) requesting re-observation. The PIs should be told that this justification will be evaluated in the framework of the original proposal. The CXC should convene a review panel consisting of a subset of the original Cycle-1 review panels which evaluated these proposals. The panel should recommend reobservations only if the main science goals were lost due to the current msec timing problems of the HRC. If an observation is redone, then the previously observed data should be made public on the date the PI receives the new data. (Note that this is the policy applied to previous reobservations due to the FEP0 problem.) 7) Cycle 2 Review of Large Proposals: The large proposals should be first reviewed in the appropriate panels and scored along with the other (small) proposals. The individual panels can recommend reduction of total time, either by reducing the number of targets or the individual time on target. However, cuts in time on target should be the exception rather than the rule. The large proposals should be selected (and targets and exposures determined) by the merging panel. 8) Cycle 2 Review of Joint Proposals: It is important to release the results of the review of the Chandra-HST proposals prior to the HST AO-10 deadline, if possible. (We recognize that this may be difficult). The CXO-HST and CXO-NOAO proposals should be reviewed mainly based on the Chandra science. The panels must give two grades: a primary grade based on the joint program, and a secondary grade based only on the Chandra science. The merging panel should assign HST or NOAO observing time to the top-rated proposals (in order of their primary grades for the joint program). Once the allotted HST or NOAO time is expended, Chandra time should be allotted to highly ranked proposals, based on their Chandra-only grade. Prior to the review the CXC staff should determine the number of HST (NOAO) orbits (nights) for each proposal and target so that this can be tracked in the electronic database. This information should be provided to the review panels. For the Cycle-3 NRA, the target forms should include a column for HST orbits or NOAO nights for each target in joint proposals. This would avoid the need to extract this information by hand. 9) TOO Observations in Joint Proposals: The Users committee was seriously disturbed by the high fraction of all of the TOO time for Cycle-2 assigned to joint HST/CXO proposals by the HST review panels. In the future, the CXC Director should exercise greater control over the allocation of TOO time by other panels. The CXC should define the joint observation budget well in advance of future cycles and consult the CUC concerning this budget. Under no circumstances should other panels be allowed to allocate TOO time in excess of the budgeted amount. 10) ACIS Squeegee Mode: The committee was impressed with the progress made in testing the ACIS "squeegee" mode to improve the CTI problem. We understand the difficulty of calibrating a large number of modes of ACIS, and therefore we concur with the proposed plan of implementing only one such mode. However, we think it is essential that the option to turn off the squeegee mode be available, particularly for grating spectroscopy. 11) ACIS Low Energy Response: The committee recommends that a high priority be accorded to determining the correct representation of the instrument performance below 0.5 keV, particularly for S3. The CXC needs to provide better and up-to-date information on such calibration issues (through the Web). 12) ACIS Spectra - PI vs. PHA: We asked for a report on the PHA-PI issue in January. The CUC renews the request for this report (to be displayed on the web). While this may be a somewhat complex issue, users need to have the best current information to analyze their data. 13) ACIS Background: We applaud the effort to provide background files and software for ACIS. The current files are screened to quiescent background. In order to help users with data with a significant portion of background flares, it would be helpful to also provide event files including events from times of high background, tagged by the background rate. 14) Facilitating Distribution of User-Developed Software and Threads: We endorse the internet "swap-meet" and other such venues to facilitate exchange of algorithms and data analysis packages. 15) Data Analysis Workshops: CXC, as the center for the premier X-ray observatory, should organize a workshop or workshops on X-ray data and analysis techniques (much in the style of NRAO workshops as well as STScI workshops). In the near term, we encourage a mini-workshop at the upcoming HEAD meeting, which will probably be most useful for people who are familiar with X-ray astronomy but need to get up to speed on the CXO instruments. A larger workshop should target students and first-time users of CXO. 16) Setting Final Observing Parameters -- Change in Policy: We appreciate the tremendous amount of time and effort that gets wasted when last minute changes are made by users in specifying the instrument parameters, or when these users repeatedly fail for respond to queries about the parameters for their observation. We concur with the proposed policy that if a PI (or a designated alternative) does not respond to a request from the USINT for final observational parameters by a well-defined deadline, then those observations can be postponed to a later date as determined by CXC staff, including a postponement to the next year. Users are of course free to appeal their case directly to the Director. It is critical that the E-mails to PIs clearly specify this policy and potential penalty. In some cases (e.g. extended sources) it would be helpful if the PI is informed of the roll angle to maximize the resulting observations. 17) Science Threads and Testing of Software: Previously, the Users Committee had recommended that CXC staff be encouraged to have active scientific programs. This request was motivated in part (other motivations included an improvement in morale) by two desires: (i) to provide extensive beta testing of software, and (ii) the development of science analysis threads. We reiterate this desire and our expectations. 18) Aspect: We agree that pixel randomization is reasonable for most programs and observers. However, we recommend that a science thread be given on the webpage to remove the randomization. This thread needs to list all parameters for the acis_process_events tool.