Notification of problem affecting certain ALMA data that used mosaic and offset pointing observing modes

The Joint ALMA Observatory (JAO) has recently identified a problem affecting selected programs from Cycles 1 to 3 where positional offsets or mosaics were used to set up Scheduling Blocks in the ALMA Observing Tool.

The problem was introduced by the ALMA online system such that there was a small mislabeling of the position of each field recorded in the data in these cases. This was due to an inconsistency between the way in which the online software computed the coordinates that are stored in the field table of the data and the way in which it computed the propagation delays and antenna pointing coordinates. It is corrected by recomputing the coordinates in the field table and the magnitude of the correction depends on the size of the offset (see Table 1 for an example). Note that the coordinates observed by the telescope are not affected.

The issue only affects programs that intended to either map extended areas around a reference (mosaics) or used offset pointings from such a reference. For such maps, the problem that has been detected would result in possible distortion of the final images that depends on the distance from the reference position, the coordinates of the reference position and the size of the area mapped (for mosaics). In addition to the normal astrometric uncertainty, the positions derived from these images will have a systematic error whose magnitude depends on the above factors. Table 1 shows an example of how the size of the effect varies at different offset distances from the reference position.

Table 1: Example of the positional offset in the ASDM field table coordinates for an observation of a source at -30 deg Declination at different offset distances from the reference position.

(1): Offset distance of the mosaic pointing or individual pointing from the reference position position in right ascension (arcsec).

(2): Positional offset of the coordinates in the ASDM field table (arcsec).

About 80 observing programs (see list below) were affected and the raw data (ASDMs) of affected Member ObsUnitSets are currently being repaired in the ALMA archive. Note that affected observing programs may have multiple datasets affected. All affected PIs will receive an email from the ALMA Helpdesk containing additional details.

For the majority of cases the effect on the data is likely to be insignificant. A subset of the programs that are more likely to be significantly affected, however, are being assigned for re-imaging. This subset of programs are marked with an asterisk in the list below.

For observing programs in this subset that are publicly available and no longer within their proprietary period, these data have been flagged in the archive as being in the process of QA3 evaluation and are no longer currently available for download. They will become available again once the QA3 re-imaging process is complete. It has been decided by the ALMA director that the proprietary period not be extended for these observing programs.

For observing programs in this subset that are still within their proprietary period, the same QA3 procedure has been implemented. In these cases, once the QA3 re-imaging process is complete PIs will be delivered the corrected data products and the proprietary period will be extended.

For observing programs that have yet to be delivered, affected data will be fixed before being sent to PIs. The normal proprietary period will apply to those data. 

The problem has now been fixed in the ALMA online software and data taken from March 16, 2016 onwards will be unaffected.

We sincerely apologize for any effect this will have on researchers utilizing these data. Additional questions, concerns or issues should be sent to the relevant ARC for support through the Archive and Data Retrieval Department of the ALMA Helpdesk. 

The affected observing programs are listed below. 

Note: (*) denotes data assigned for reimaging.