Maintained by CGWW
Maintained by CGWW
2010 June 9
PS1 is taking science quality data on a nightly basis, and since new moon on May 13 has been sustaining the Modified Design Reference Mission at an acceptable level. Data reduction and distribution are largely automated. There are ongoing minor issues in several subsystems, but the PS1 Science Mission is effectively underway, and modulo modest variations from the MDRM working group, it should be sustainable for the foreseeable future.
This will be announced shortly by coordinated press releases from each PS1 Science Consortium institution.
The black covers on the primary mirror support cell have had a significant impact in reducing scattered light that contributes to diffuse sky background. Formal analysis of results measurements are underway, but this was clearly a successful effort. Further scattered light mitigation efforts are planned.
Instrument rotator has intermittent failures. Currently is at most only once every few nights and it takes two or three minutes to recover, and if the failure was during an exposure, it requires a repeat of that exposure. Some suspicious connectors and components have been replaced, but the problem has not been eliminated. This would be at the nuisance level except for the possibility that the problem is in the rotator motor / harmonic drive assembly itself. A test stand for a spare rotator motor assembly is under development and will allow daytime tests to determine if the problem is in the (accessible) electronics or the much more difficult to replace motor assembly itself. But at present the failure rate is not having a significant impact on observing.
A temporary fix has solved the dome cable wrap limit switch problem, with a more robust fix planned. Other issues with the dome drives are under investigation, but we are no longer getting failures that require a trip to the summit.
The Project has provided a new focus model based on data taken since the work on the primary and secondary support systems, and it includes new temperature sensors. The application of the qtfocus algorithm still needs some tuning, and this is under investigation.
Software work has improved the camera overheads by fixing the size of data transferred for psf analysis and focus. Further small improvements are expected. Several "microtest" suites of alternate camera settings have been tested on the sky, and tuning of individual OTA's could potentially reduce the number of pixels currently being masked by the IPP.
Otis is operational and has rudimentary tools for scheduling the current version of the MDRM. The Project is making progress on the deliverables, which should improve the efficiency and usability of the scheduling tools and of reporting of observing statistics and overheads. A significant change from the MDRM working group in the observing strategy will have an unfortunate but necessary impact on the progress of the deliverables. The most immediate is a new 3pi tessellation, which will be in place Friday on new moon, and thanks to the Project for making the implementation of this a priority. The new tessellation was unanimously endorsed by the MDRM working group, and will reduce the number of pointings in the 3pi survey to 4412, giving sufficient margin to ensure 3pi survey completion with current exposure times. Further improvements in operational overheads are expected, and when in place, the "Calibration Fields" will be added back into the observing schedule.
The IPP team has been finishing the latest release with all the accumulated bug fixes and new code, including an approved version of "magic." This version has all the forced aperture photometry capability and the automated stack photometry. The automation is now complete, and a static sky analysis can be commanded with any set of input stacks or filters. The night to night processing has kept up, but the release of the DVO catalogs of 3pi has fallen behind as the team has been preparing for the code release. A stack of all the historical MD04 data (100-300 images per filter) is being generated and will be examined in detail by the DRAVG. There is no known optimal stacking algorithm, the issue is whether it is "good enough" operationally - the handling of the complete covariance matrix has not been done before in any pipeline system.
PSPS is functional and has ingested all the "Demo Month" data. It is now ready for consortium scientist "Beta Testers" to start using it Monday (June 7).
Data Release Outlook:
There is a confluence of events that occurs near the first of September:
1. The "edge effects" of starting the survey will be over, and we will have "first year" 5 band stacked
3pi data coming in at the sidereal rate (4 exposures in each filter).
2. The total PS1 commissioning data to date form a nearly complete (in one band or another) astrometric catalog that will be an improvement over anything available for registration and warping of 3pi data.
3. Additional "wave 4" IPP hardware will be installed and running.
4. The PSPS should be tested and ready for operations.
So the plan is to have in place and functional
1. a stable/frozen production release of the IPP code;
2. automated processing and distribution of all new data with less than 24 hour latency, possibly approaching the long-term design latency of 30 minutes;
3. start the re-reduction in parallel of all data prior to Sept 1;
4. Flush the "beta test" of PSPS and start the weekly ingestion of reduced IPP data products into PSPS.
We would also like to start formal integration and operations of those science servers that are ready to produce derived data products, and advance the state of development of the outstanding ones.