Using the advance notice of pointing


Part of the specification of the DM interface with the Observatory Control System (OCS) is that we are given advance notice of each pointing - 20 seconds ahead of the start of the first exposure of a standard visit. (OCS-DM-COM-ICD-0031 in LSE-72) This notice includes a boresight (ra,dec) and a rotation angle. (The azimuth and elevation would also be available, probably referenced to the time of the start of the first exposure.)

The purpose of this advance notice is to give the Level 1 system time to retrieve the template and other reference data for the Alert Production analysis of the visit, on the theory that this work can usefully be overlapped with the processing of previous visit(s).

If the actual pointing is different from the advance notice, or the advance notice is not received, the Level 1 processing is still required to produce the same quality of processing of the visit that it normally would, but it is relieved of the usual 60 second latency requirement. That is: the advance notice serves only as a performance optimization.


Is there a use case in DM for using this nominal pointing - or, for that matter, the actual pointing from the OCS/TCS after the slew (currently not specifically called out as required by DM) - to generate a sort of “pseudo-WCS” for the visit? That is, an estimate of what the WCS might be if the pointing were correct?

More generally, I’ve been asked by a Telescope system colleague: do/will we have an API for generating such an object from a nominal pointing (and the camera geometry)? This would be useful for, for instance, estimating the expected position of a reference star on a guide sensor.

Isn’t this pseudo-WCS something that palpy or equivalent would produce essentially trivially?

We do not currently have an API that does this directly, but I think we do have the necessary interfaces in camera geometry to implement this in a few lines a camera-generic way.