Muthusubramanian, Balaji (2019). Simulations of group-delay tracking strategies in the astronomical K and L bands. For improved sensitivities at the Very Large Telescope Interferometer. PhD thesis, Universität zu Köln.

[img]
Preview
PDF
main.pdf - Accepted Version

Download (8MB) | Preview

Abstract

Motivation: Atmospheric turbulence is a major issue in achieving high angular resolutions for optical and infrared astronomical observations. Such turbulences induce phase fluctuations on the incoming stellar wavefronts. With the recent advances in Adaptive Optics, we can overcome this effect and remove most of the aberrations on the stellar wavefront. However, the global phase shift (piston) which is not corrected by the adaptive optics system, plays a significant role in optical and infrared interferometry. The piston or Optical Path Difference (OPD) fluctuations induced by atmospheric turbulence and instrumental vibrations at an interferometer affects the position of the fringes on the detector and thus influencing the visibility measurements. Thus, a dedicated fringe tracker is required in addition to the adaptive optics system in an optical or infrared interferometer. The fringes can be tracked in two ways, namely group delay tracking and phase delay tracking. Phase delay tracking involves in tracking each individual fringe phases and group delay tracking involves in tracking the fringe envelope. Currently the Very Large Telescope Interferometer (VLTI) houses GRAVITY fringe tracker, FINITO and PRIMA FSU which tracks group and phase delays in H and K bands. However, targets such as Young Stellar Objects (YSO), dusty torus around Active Galactic Nuclei (AGN), etc. are brighter at longer wavelengths due to the presence of dust around them. Hence in this work, I explore the possibility of on-source fringe tracking in the L band alongside K band. Fringe tracking in L band is advantageous due to its longer atmospheric coherence time and larger Fried parameter as compared to that of shorter wavelengths. I also extend my work in performing feed-forward group delay estimation at mid-infrared wavelengths using inter-band dispersion between K and L bands. The main goal of the work is to analyze the performance of a simultaneous K and L band group delay tracker which keeps the fringe frames of the science channel within coherence length. The sensitivity of the science channel can then be increased by stacking each frame by its power spectrum. Method: To analyze the performance of a fringe tracker at K and L bands, I have developed a atmospheric turbulence simulator (GDsim) which includes atmospheric water vapor dispersion and instrumental vibrations at the VLTI. The simulated OPD fluctuations are then injected into simulated noisy K and L band fringes over different lengths integration times. Group delay is then estimated from the resulting fringes through spectrally dispersed ABCD encoding technique, least squares estimation and the double Fourier technique. The performance and sensitivity of each of the group delay estimation algorithm is then estimated by comparing the simulated input group delay and the measured group delay. Result: The loss in fringe contrast over integration time follows a random walk behavior due to the temporal decorrelation of the phase fluctuations at each telescope over long integration times (Tint > 1.79�0). Thus the Signal to Noise Ratio (SNR) on the correlated flux is unreliable in estimating the sensitivity of a group delay tracker at such integration times. Alternatively, by analyzing the Root Mean Squared (RMS) on the residual group delay for each group delay estimation algorithm, we can estimate the sensitivity and performance of the group delay tracker simultaneously. However, the threshold on the RMS of the residual group delay (input group delay - measured group delay) depends on the spectral resolution of the science channel for incoherent stacking of fringe frames. For astronomical observations in the L band with MATISSE at low spectral resolution (R=30) with an acceptable maximum loss in fringe contrast of 10%, we can perform coherencing with a K and L band sensitivities in the range of 13.36 - 14.67 magnitudes and 6.64 - 10.02 magnitudes respectively for the different group delay estimation algorithms. Feed forward group delay tracking performance and sensitivity at mid-infrared wavelengths (M & N bands) can be estimated by comparing the individual group delay estimation errors in K and L bands. The sensitivity to perform feed forward group delay tracking using only K and L band group delay information for science in the N band with MATISSE at low spectral resolution (R=30) is 9.0 magnitude in the K band and 8.3 magnitude in the L band. Also, only the spectrally dispersed ABCD encoding algorithm in K and L bands can coherence the N band fringes with a maximum loss in fringe contrast of 10%. However, this sensitivity applies only for instantaneous estimation of water vapor dispersion.The feed forward sensitivity can be enhanced by averaging the water vapor content over multiple duty cycles of K and L band group delay estimation. The L band group delay estimation sensitivity can also be enhanced by coherencing the L band fringe tracking channel using K band group delay information at a faster duty cycle.

Item Type: Thesis (PhD thesis)
Creators:
CreatorsEmailORCIDORCID Put Code
Muthusubramanian, Balajimuthusu@ph1.uni-koeln.deUNSPECIFIEDUNSPECIFIED
URN: urn:nbn:de:hbz:38-95417
Date: 2019
Language: English
Faculty: Faculty of Mathematics and Natural Sciences
Divisions: Faculty of Mathematics and Natural Sciences > Department of Physics > Institute of Physics I
Subjects: Natural sciences and mathematics
Physics
Uncontrolled Keywords:
KeywordsLanguage
Fringe TrackingEnglish
InterferometryEnglish
Integrated OpticsEnglish
Very Large Telescope InterferometerEnglish
Young Stellar ObjectsEnglish
Date of oral exam: 23 April 2018
Referee:
NameAcademic Title
Labadie, LucasProf. Dr.
Refereed: Yes
URI: http://kups.ub.uni-koeln.de/id/eprint/9541

Downloads

Downloads per month over past year

Export

Actions (login required)

View Item View Item