Abstract:
Modern astronomical spectrographs have been in use for several hundred
years now, contributing to the study of the physical, chemical and kinematical
properties of stars and other celestial bodies. In the recent years, there has
been new developement in the design choices, calibration sources and detec-
tors for spectrograph that led to large improvements in resolution, efficiency
and stability, which have been a key to the discoveries of low mass exoplanets,
cosmic dynamics and variability of physical constants. Even with these major
improvements in the instrumentation, a little has changed in the way data is
calibrated. The physics used in the designing and building of the instrument
is seldom put to use for instrument calibrations.
In the current work, we have developed a computing scheme whereby an
astronomical spectrograph can be modelled and controlled without recourse
to a ray-tracing program. This is achieved by using paraxial ray tracing, exact
corrections for certain surface types and the aberration coefficients of Buch-
dahl for more complex modules. We have shown that the resultant chain of
paraxial ray trace matrices and correction matrices can predict the location of
any spectral line on the detector under all normal operating conditions with
a high degree of certainty. This kind of model will allow a semi-autonomous
control via simple in-house, program modules. The theory and formulation
in developing a model for HESP (Hanle Echelle Spectrograph), comparison
with commercial ray tracing software and validation the model with calibra-
tion exposures taken in the lab are described in detail. We have demonstrated
that the wavelength calibration using the dispersion solution from model per-
forms superior to the emipirical solution that is used in standard spectroscopic
reduction software (e.g. IRAF), especially in the regime of sparse and low
signal calibration lines. We also discuss the potential use of model in si-
multaneous calibration using double fiber in tracking the instrument drifts
and issues related to positional and wavelength dependancy of instrumental
drift. We conclude by listing several science areas that would benefit using a
model based calibration. Apart from calibration of science data, the model
along with an optimization routine can be used in the instrument mainte-
nance pipeline to study the real time behaviour of the instrument, trending
and long-term stability. We also present the work related to the development
of autoguiding alogrithm and fiber scrambling for HESP and implication to
radial velocity shift.