Please note that when delivered to the observer, this file also contains the XMM SSC data screening report for the observation, appended at the end. 1 PCMS Processing ----------------- Full release note for prod configuration 01_prod8.3_20100920.120500 =================================================================== PCMS Infrastructure Version: 8.3 Pipeline Version: 01_prod8.3_20100920.120500 Pipeline Name: prod8.3 Release Date: 2010-09-20 SAS Manifest: xmmsas_20100920_1801 SAS Version: 10.0.3 ppstrack Summary ======= This pipeline represents a significant change with respect to the previous prod8.1 pipeline (version:07000001_20090303.144400), embodying both a substantial number of (mostly minor) changes to the PCMS processing system itself and incorporation of version 10.0.3 of the SAS. The most significant changes to the pipeline include: i) Creation of per-band detector masks. In the previous pipeline, PN bad (dark) columns in the band 1 and 2 exposure maps were based on band 1 information while band 3-5 exposure maps were based on band 3. However, for the background maps, dark columns were mapped based on band 1 detector masks only. As a result, PN band 3-5 exposure maps differed in appearance from the background maps in terms of the dark columns imprinted on them. This caused no material difference in source detection as the (conservative) band-1 detector mask is applied but it led to visual differences in the band 3-5 exposure and background maps. In this new pipeline, this discrepancy in the exposure/background maps is removed by the creation and use of the multi-band detector masks. Note that the band-1 detector mask is, however, still applied, as before, in source detection. ii) Inclusion, for the first time, of the epreject task to better filter PN low energy MIP events Other upgrades involve: iii) A number of minor bugs and cosmetic improvements have been made to the PPS html output files. PN SW and MOS W6 modes are now (correctly) shown as not used for source-specific-product creation. The precision of some quantities (e.g. coordinates, detection likelihood) are made more consistent, and the coordinate error units are now (correctly) stated as arcseconds. Rounding errors on coordinates should no longer yield 60.0 for the seconds/arcseconds element. iv) EPIC flare background timeseries graphic plots are now plotted with the count rate (y axis) in log space to better highlight the low rate periods. v) In cases where the OPEN filter is used with the EPIC instruments, these data are now excluded from the source detection stage to reduce the potential impact of optically loaded sources. The OPEN filter is only used extremely rarely. vi) Where the variability test was not applied or failed during the processing of EPIC source specific time-series products, the AVRATE, CHISQUAR and CHI2PROB keywords are set to -1 and N_POINTS=0. vii) Systematic errors are now set to the values of 0.35 (where astrometric rectification was successful) and 1.0 otherwise, as explained in Watson et al., 2009, A&A, 493, 339-373. These values are propagated for use by the ACDS external catalogue cross-correlation stage. viii) The USNO B-1 extract used during the astrometric rectification of both EPIC and OM source lists is created as a formal product for the first time. This file includes USNO B-1 objects in the full EPIC field. ix) The nsigma parameter is raised from 2 to 3 for omdetect in OM fast mode processing. The changes of the SAS between SAS version 8.0 used in pipeline 07000001_20090303.144400 and SAS version 10.0.2 are described at http://xmm.esac.esa.int/sas/current/documentation/releasenotes. Minor further changes were included in to SAS 10.0.3 which is used in this pipeline. Documentation ============= Documentation for this pipeline and pipeline products can be found here: