Use Case: ComputePointingModel


The Calibration Pipeline use stored results of previous pointing measurements to compute a global or local pointing model in an automatic way. Local pointing model must be computed in quasi-real time.

Role(s)/Actor(s):
Primary: Calibration Pipeline
Secondary:

Priority: ***

Performance: < seconds.

Frequency: minutes to hour.

Preconditions:

  1. The session processing is initiated by the Calibration Pipeline when all results of pointing measurements on different calibrators are available.
  2. Results with a non NULL error code have been filtered out.
  3. Input1: per source and antenna
    • Azimuth collimation correction and associated uncertainty per antenna;
    • Elevation correction and associated uncertainty per antenna
      • ;
    • Quality flag.
  4. Input2: per antenna
    • Default pointing model parameters.
  5. Input3:
    • Kind of pointing model asked.

Basic Course:

  1. Verification of input data consistency (e.g. does quality flags are satisfactory? Does the number of successfully measured sources enough for the kind of asked model? Does the successfully measured sources well scattered in azimuth and elevation for the kind of asked model?).

    2. Exception Course: Processing is stopped and return a specific error code.
  2. The antenna-based pointing models are computed together with uncertainties.

    3. Exception Course: Processing is stopped and return a specific error code.
  3. Verify whether solution is sensible (e.g. compare new model parameters with old ones) and return a quality flag for each model parameter (some parameters are harder to find than others) and/or an alarm when corrections are large (as the model parameters should evolves smoothly if no modifications have been done on antennas).

    4. Exception Course: Processing is stopped and return a specific error code.


Alternate Course: None.


Postconditions:

  1. Output:
    • New pointing model parameters (To be filled) and associated uncertainties;
    • Quality flag;
    • Error code.
  2. The output is fed back to the Calibration Pipeline.


Issues to be Determined or Resolved:

  1. There are uncertainties about the fact that local pointing models will often be used although this is the assumption here. The local pointing model characteristics (validity in sky area and time, number of source sneeded) are a current subject of research in the Science IPT.

Notes:

  1. ***

Owner: Jerome Pety.
Last modified by $Author: lucas $ on $Date: 2003/02/14 12:13:08 $