HARPS Support Software

Exposure Time Calculator

An Exposure Time Calculator (ETC) is available. It is modelled after the UVES ETC but uses the efficiencies relevant for HARPS (atmosphere, Cassegrain telescope, seeing, "slit loss", instrument optics (including fibre link), CCD). During the Commissioning 1+2 runs the ETC proved to be accurate to within about 10%. After the overhaul of the secondary mirror of the 3.6m telescope the instrument efficiency increased by about 15%. This is not taken into account yet by the ETC. When using the ETC for the Iodine self-calibration method then one must in addition take into account the ~50% absorption loss in the Iodine cell which the ETC does not include in the calculations.


This is the standard ESO way of operating an instrument: OBs are prepared from instrument-specific templates with the P2PP tool before the observations. At the telescope, OBs selected by P2PP are fetched by BOB and executed. Details about P2PP are found here. Althought the P2PP manual might state differently, please, be aware the the proper motion units which are needed by the system are arcsec/year.

Data Reduction Pipeline

The HARPS Consortium has developed an extremely sophisticated piece of software which allows reduction of the observed spectra in all observing modes in near real-time. The readout time of the instrument is about the same than the pipeline processing time for scientific observations, this guarantees that the pipeline reduction will never lack behind. The reduction depends on the observing mode and on the spectral type of the star. These parameters are retrieved by the pipeline from the FITS header of the raw data. For proper reduction to achieve the highest radial velocity accuracy it is crucial to run the standard afternoon calibration set (see the calibration page). Apart from the online pipeline, which is started in the afternoon by the telescope operator, the user has available an offline software which allows to reprocess the data. This is especially useful for objects for which the radial velocity is not known. In this case a first attempt to determine the radial velocity could be done by using a large window for the correlation to be performed. On a second stage the radial velocity computation can be re-done using the previously calculated value as a first guess value.
For optimum radial velocity calculation (1m/s) the radial velocity of the object has to be known within 2 km/s. The offline pipeline can be started by the user via the command offdrs.csh from the data reduction machine (whaldrs) at the telescope consolle. Its usage is self explanatory. Several recepies can be used, from sinple data visualization (visu) to cross correlation function (and RV) recalculation (new_ccf). The newly calculated cross correlation will be stored together with all other pipeline processed data and the filename will have the prefix new_ccf.

Results of the reduction are:

  • dark current values, order localization, flat fields, dispersion solution (for calibrations);
  • extracted spectrum (all modes)
  • precise radial velocity (only simultaneous Th reference);
  • cross correlation function (only simultaneous Th reference).

The pipeline is running automatically for every spectrum at the telescope. No intervention is required to the user, apart acknoledging the eventual warning pop up windows, in which case it is recommended to contact the support astornomer. Further details about the Data Reduction Software (DRS) can be found in the DRS manual.

The release history of the pipeline is in this page.
You're on EPU
View live page