Viking Orbiter Image Times Summary =========================================================================== By Boris Semenov and Chuck Acton, NAIF/JPL, July 22, 1996; updated December 3, 2008. Introduction -------------------------------------------------------- Two SPICE E-Kernel files, ``vo1_time.bdb'' and ``vo2_time.bdb'', provided in this data set contain all variants of VO1 and VO2 time tags that have been collected by NAIF from science groups involved in VO image cartographic application. Time tag information from the following data sets has been included in these two DBK files: -- times of the images used by R.M. Batson of the USGS Astrogeology Branch (BATSON set); -- times of the images used by M. Davies of the Rand Corporation (DAVIES set); -- times of the images used in GEM (Viking Image Geometry) files produced by K.Edwards of the USGS (GEM set); -- times of the images used in the USGS Mars Digital Image Mosaic (MOSAIC set); -- times of the images used by Malin Space Science Systems company (M.Malin, M.Caplinger) for automatic and manual image registration (MSSS_AUTO and MSSS_MANUAL sets); -- times used in the original VO SEDR data set generated by the Viking Project (SEDR set); -- times contained in the PDS Viking image catalog at Washington University, Saint Louis (WASHU set). The times in a few of these sets (BATSON, WU, MOSAIC, both MSSS sets and part of DAVIES set) are mostly the same as those in the GEM set. But since during data processing some errors were introduced especially in the BATSON set) each of these data sets are included in this DBK file. Moreover, since not all VO images were used in the cartographic application of each group, all ``derived'' sets contain significantly smaller numbers of time tags. Perhaps only the three most complete sets will be of an interest for a future VO image analysis; these are SEDR, GEM and WASHU time tag sets. Since the SEDR set is the only set produced during the mission, it is assumed that the GEM and WASHU time tags were ``derived'' from it later during VO data processing. Known and possible origins of time tag differences -------------------------------------------------------- Unfortunately the time tags in these three sets are different. It's very difficult to decide now, many years after the mission, what are the origins of these differences. One of the main differences stems from the way time tags were stored in each of these sets. In the GEM files tags were stores in are ``middle-of-exposure'' times in JED date format (ex. 2443365.317484 for FSC 33919977). All these tags are truncated to 6 decimals. This causes a loss of precision since .000001 of day corresponds to 0.0864 seconds. The time tags stored in SEDR files are middle-of-exposure times in UTC day-of-year format (ex. 1977-221 // 19:36:22.570 for FSC 33919977). Here the tags are given to a millisecond. The tags in WASHU set (PDS catalog) are beginning-of-exposure time (in accordance to Ed Guinness) in UTC Calendar date format (ex. 1977-08-09T19:36:22Z for FSC 33919977). These are given to a seconds. For the period of time starting at Jan. 1, 1978 and ending about three months later differences between SEDR and GEM and SEDR and WASHU time tags are greater that 1 second. This can be explained with the assumption made above that GEM and WASHU tags were derived from SEDR tags. Then possibly during this process strange looking SEDR UTC DOY-format tags like ``1977-375//...'' (instead of ``1978-10/...'') were converted to ET incorrectly. i.e. the leap second of Jan 1, 1978 was not taken into account when derived time tags were calculated. To make comparisons of tags in these sets easier the differences between the SEDR tags and the corresponding tags from each of other sets are provided in separate columns in this EK file. -------------------------------------------------------- Statistics and specific problems resulting from a comparison of these data sets are in several tables at the end of this document. Recommendations -------------------------------------------------------- What recommendations can be given to a person who will process VO SPICE data in the future? There is not an obvious answer, but a good bet will be to use original SEDR time tags. Data set comparison statistics -------------------------------------------------------- TABLE 1. Number of records in VO1 time tag data sets in source included files in DBK ---------- --------- BATSON 9337 9328 (-7 dupl, -2 bad FSCs) DAVIES 1315 1315 GEM 35569 35569 MOSAIC 3521 3521 MSSS_AUTO 3086 3086 MSSS_MAN 8175 8175 SEDR 33517 33517 WASHU 34918 34918 SWU 587 587 Some FSCs were duplicated in VO1 BATSON data set; i.e. the same FSC value was included twice with different corresponding JED times. For each pair of duplicate FSCs the FSC whose corresponding JED time seems to be incorrect was not included in this DBK file. These ``not included'' FSC-JED pair are below: 26923395 2443156.5513080000 34700402 2443405.7840340000 39256001 2443642.0002380000 39294549 2443643.9991270000 39294552 2443643.9991790000 52513428 2443038.6449790000 54395405 2444427.0070470000 A few FSC tags found in the BATSON data set couldn't be found in any of the most complete data sets -- SEDR, GEM or WASHU. These FSCs were also not included in this DBK file but are listed below: 43485635 2443861.3157300000 54423249 2443334.3385800000 TABLE 2. Number of records in VO2 time tag data sets in source included files in DBK ---------- --------- BATSON 2431 2430 (-1 duplicate FSC) DAVIES 328 327 (-1 not found FSC) GEM 16132 16132 MOSAIC 1060 1060 MSSS_AUTO 459 459 MSSS_MANUAL 2415 2415 SEDR 15621 15621 WASHU 15582 15582 SWU 125 125 One FSC was duplicated in the VO2 BATSON data set; i.e. the same FSC value was included twice with different corresponding JED times. For this pair of duplicate FSCs the FSC whose corresponding JED time seems to be incorrect was not included in this DBK file. This ``not included'' FSC-JED pair is below: 56275610 2443541.7246200000 Also one VO2 FSC found in the DAVIES data set couldn't be found in any of the most complete data sets -- SEDR, GEM or WASHU. It was not included in this DBK since the JED time corresponding to it seem to be incorrect: 49792705 2443169.6267499800