pmtras_database.txt	2002-jul-05

PMTRAS solution database


RATIONALE

It is much easier to tease a good solution out of the data for an optimized time period than
to make the software have the ability to autonomously get a good solution for a user-selected time interval.

Cannot expect users to tease solution out of the data.

Data base of roll solutions is likely to be very stable.

The last straw:  The discovery of daily gaps in pmtras data.

Another consideration:	Needed for applications using very long time intervals.

Opens up more options for interpolating into data gaps.  (eg combining pre and post
gap solutions and orbital patterns...)

RAS data to date may not have the required accuracy (due to lack of time-interpolation).


APPROACH:

Create a data base consisting of a table of anytime, pa, quality flag (8+8+4 bytes), tabulated
once per minute.  (==>30 kbytes/day, 10 Mbytes/year)

Data base could be distributed as part of SSW.

To generate this, would use a batch program to run pmtras_analysis one orbit at a time,
generating individual output files and a quality database.  Then could rerun this interactively
for the problem orbits.  A separate merge program would combine individual outputs into a single
data base.  Time periods would overlap slightly so tht the merge program could confirm continuity of
roll solution.  Merge program would also modify the solutions so that pa increased monatonically
except for occasional short (rewind) discontinuities.  Merge program would also generate a data base
of roll periods and other measures.

The use of the tabulated data base would be the default for imaging, but users would always
have the option of doing their own solution.

ISSUE:  How to coordinate this with the need for quicklook flare locations.

------------------------------------


DATABASE GENERATION:

PMTRAS_DATABASE_GENERATOR uses multiple calls to pmtras_analysis, one orbit at a time, to
		generate a set of one-orbit solutions.  Output is combined into one quality-flagged
		output solution file.
		Secondary outputs:  	1. List of failures
					2. List of gaps
					3. Appends output to a data base of star intensities and 
									polar-angle systematics

PMTRAS_DATABASE_EDITOR  Edits a specified set of orbital files into existing data base
		after checking for continuity.  Also arranges for the monatonic increase of roll angle.
		Secondary outputs:	1. List of gaps
					2. List of discontinuities
													
PMTRAS_GAP_FILLER uses the interim data base (+ partial pmtras solution ?) to estimate
		roll solution in the gaps.  Generates file in same format as pmtras_analysis for
		use as input to PMTRAS_DATABASE_EDITOR.	NB.  Tolerance for gaps should depend on
		whether it is in daylight or SAA.