See the Final 2MASS Data Processing Scope document for an overview of the scope of final processing.
Process only calibration scans that were found to define photometric periods in preliminary processing quality assurance. No attempt will be made to recover time that was previously ranked to be non-photometric, but rather to obtain the best measure of the photometric zero point offset at any time during the night.
This implies that calibration scan sets rejected in overrides in the
nightly QA process for reasons such as poor seeing, etc. may be excluded
from processing and excluded in the final fits by default.
Process only science scans that were assigned a quality score Q>3 during preliminary processing quality assurance. No attempt will be made to revive scans that were previously rejected as non-photometric in periods between photometric intervals that may now be fit in the photometric solutions through the use of higher order zero point curve fitting.
This implies that survey scans rejected for any reason, such as insects on the window and bright aircraft fly-bys will not be recovered.
As with preliminary processing, final processing will be performed on all data obtained on individual nights. For each observatory, the nights will be processed in the order in which they were acquired. Two parallel processing streams will be run, one running northern data and one running southern data.
The purposes of quality assurance review during final 2MASS processing are to validate QA scoring during preliminary processing and to apply uniform QA ranking to all data (some of the review parameters and algorithms were changed during the course of the survey). As such, and because of the schedule constraints on final processing, QA review for final processing will not be as extensive and human-intensive as during the preliminary processing.
Final processing quality assurance will evaluate the same suite of
QA metrics and summaries evaluated for each scan as during preliminary
processing. These new metrics should be compared with the old QA results
and trigger alerts when they differ from the old values by specified values.
The purpose of photometric calibration in final processing will be to evaluate the best measure of the photometric zero point offset as a function of time during the night. We will not attempt to extend the nominal photometric intervals for the purpose of recovering survey scans previously scored as non-phometric.
Recognizing that the photometric calibration from most nights during the
survey was well-behaved, calibration quality assurance should require
close scrutiny beyond a visual inspection of the zero point plot only if
the zero point correction applied to any survey scan exceeds a TBD value,
or if the resulting RMS to the fit exceeds a TBD value.
Each survey scan will be assigned a final quality score value between 0 and 10, with 10 denoting the highest quality data.
All output of the final processing will be archived to standard processing archive tapes. A subset of output diagnostics and statistical summaries will be archived on-line in the production archive (PARC) directories on server lippi. Source lists and scan information summaries for all survey scans assigned a quality score Q>5 (TBD) will be archived in Working Databases in the Infrared Science Archive (IRSA) system. Lossy-compressed images from Q>5 (TBD) survey scans will be archived on-line on the lippi image server. Non-compressed Atlas Images from from Q>5 (TBD) survey scans will be archived on tape, and on the HPSS tape robot at the San Diego Supercomputing Center.
Issues: