We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Centre for Advanced Instrumentation


CANARY is an open-loop Adaptive Optics system that will be deployed on the 4.2m William Herschel Telescope in La Palma from July 2010. CANARY will use multiple laser guide stars and deformable mirrors and will be the instrument that provides the first on-sky test of combined wide-field LGS tomography and open-loop AO control. On-sky validation of these combined techniques will be performed with the goal of emulating a single channel of the proposed E-ELT MOAO instrument, EAGLE, albeit at 1/10th scale. The EAGLE webpages contain many examples of the type of science that could be performed with such an MOAO instrument on the E-ELT.

CANARY Development

Building and testing an MOAO instrument with the same configuration as EAGLE, with multiple DMs and both natural and artificial (Laser) reference stars is undoubtedly complex. In order to reduce the overall difficulty of building and testing such a system, the CANARY team have adopted a phased approach to CANARY development with on-sky runs over 5 years from July 2010.

  • Phase A (2010): 3 x NGS WFS tomography and a single low-order DM
  • Phase B (2012): 4 x LGS WFS tomography and a single low-order DM
  • Phase C (2013-2014): 4 x LGS WFS tomography and a woofer-tweeter DM configuration

CANARY design

In order to allow the 3 CANARY phases to be conducted with only minor modifications and upgrades required, the CANARY design is based around a set of reconfigurable optical modules. By re-arranging these modules, different instrumental configurations can be achieved for the three phases without redesigning the entire system.

Fig 1. Layout of principal opto-mechanical modules at Phase A

The principal optomechanical modules are shown in their phase A (3 x NGS) configuration in the figure above. The optical layout of the main AO path modules is shown below.

Fig 2. CANARY phase A optical design. The light from the WHT enters from the left of the image

The WFSs and DMs are linked together by the Real Time Control System (RTCS) which links to the rest of CANARY through the Instrument Control System (ICS). You can read more about both the RTCS and ICS under the 'Projects' heading.

CANARY Performance

Monte-Carlo simulations of the CANARY system were performed at Durham and at the Observatiore de Paris to determine the level of correction that could be achieved using standard asterisms and representative turbulence profiles for La Palma. For a 3-star triangular NGS asterism (30" radius) CANARY will achieve a Strehl ratio of approximately 0.22 at a wavelength of 1.65µm (323nm RMS error). The major source of error within this system is due to the sampling of the highest layers of turbulence at an altitude of approximately 15km above the telescope. With the defined asterism on a 4.2m diameter telescope, the distance between two of the projected pupils at 15km is approximately 3.8m. If the pupils do not overlap, then no tomographic information can be derived from this layer thereby degrading performance. Other significant sources of error within the system are the DM fitting and temporal errors.

In order to measure and characterise all error sources within CANARY the system contains many calibration and alignment tools that would not normally be included with a closed-loop facility-class instrument. One of the most critical components are the turbulence profilers that are able to characterise the vertical distribution of turbulence above the telescope. In addition to the WHT's own turbulence profilers, two SLODAR systems will be used. The first SLODAR system will use the information from the CANARY WFSs, while a second external SLODAR system will be installed on the WHT roof to measure the turbulence up to 20km.


Unlike most instruments deployed on large astronomical telescopes, the name CANARY isn't an acronym for anything. The CANARY name harks back to the first proposed MOAO instrument for one of the 8m-diameter VLTs, called FALCON. The final phase of CANARY is also demonstrating a single MOAO channel for the proposed EAGLE instrument on the future 42m diameter European ELT.

The avian-related naming scheme for MOAO instruments has recently been extended to include RAVEN, an MOAO instrument being developed by HIA. Other on-sky MOAO instruments include ViLLAGES and VOLT.

CANARY is a collaborative project between several institutions in the UK, France and Spain (shown below) and is funded by STFC (under the UK E-ELT Design Study), ANR Mauii, INSU, Observatiore de Paris, Durham University and the EU under FP7 Preparatory Fund WP9000 and FP7 OPTICON JRA-1

Status updates

Current Status (23rd August 2016)

The latest upgrade to the CANARY adaptive optics system was commissioned between the 19th and 25th July with the installation of the European Southern Observatory’s ‘Wendelstein’ sodium Laser Guide Star Unit (WLGSU) in a new custom-built laser enclosure next to the 4.2m William Herschel Telescope in La Palma.

The 20 Watt 589nm laser was used to excite fluorescence of a thick layer of sodium atoms which are typically observed between 80 and 100km above sea level. The wavefront distortions caused by the turbulent atmosphere in the returned laser light were measured using CANARY 150 times a second and will be compared to the distortions measured from a constellation of background stars.

Unlike other LGS systems in operation, the laser is situated 40m off-axis from the WHT, recreating the LGS geometry expected for the upcoming European Extremely Large Telescope ( and turning the WHT into a 4.2m diameter segment of this giant telescope. Recreating this geometry is important because the variable thickness and density of the sodium layer means that the LGS suffers from 20-30 arcseconds of perspective elongation in the CANARY wavefront sensors. LGS on existing 8-10m class telescopes typically observe 2-5 arcseconds of elongation, which is inside the isoplanatic patch size of the atmosphere. The first generation of instruments proposed for the E-ELT all rely on adaptive optics system to fulfil their scientific goals, and errors in the wavefront sensing coming from the highly-elongated LGS will affect performance. There are several proposed methods for wavefront sensing from such an elongated LGS, but these have never been investigated on-sky in real-world environment.

To ensure a measurement of the LGS wavefront that can be disentangled from the effects of density variations in the sodium layer the 2.5m INT is also being used to observe the LGS. The INT is 420m off-axis from the LGS launch location resulting in elongations of up to 5 arcminutes. A 6.5 arcminute field of view laser plume imager running at 150Hz (synchronised with the CANARY wavefront sensors) was used to provide a high vertical resolution image of the plume that can be used to calibrate the CANARY wavefront sensor measurements. The INT telescope control system was integrated with the CANARY systems to allow pointing, focus and rotation commands to be offloaded automatically from either CANARY or the plume imager, turning the INT into a robotic (albeit monitored!) telescope. Simultaneous measurements of LGS return flux and atmospheric transmission are also provided by the WLGSU via an auxiliary 35cm diameter telescope installed next to the laser enclosure.

The CANARY AO demonstrator at the WHT with ESO’s WLGSU is the only facility worldwide capable of performing a comprehensive field investigation of the effects of extreme LGS elongation on AO performance. Results and analysis from CANARY will feed directly into the next generation of ELT scale AO instrumentation.

CANARY Phase D is an international collaboration consisting of Durham University, Observatoire de Paris, European Southern Observatory, INAF - Osservatorio Astronomico di Roma, Instituto de Astrofisica de Canarias, UK Astronomy Technology Centre and the Isaac Newton Group of Telescopes. CANARY is funded in the UK by STFC through the UK E-ELT program (Ref ST/M007669/1), by CNRS/INSU and Observatoire de Paris in France and by the European Commission OPTICON project (EC FP7 grant agreement 312430). The WLGSU is funded by ESO. The CANARY and LGS teams would like to thank the ING and IAC for their support before and during the run.

More information:

“Proposal for a field experiment of elongated Na LGS wave-front sensing in the perspective of the E-ELT”, G. Rousset et al, Proc. SPIE 9148, Adaptive Optics Systems IV, 91483M (2014); doi:10.1117/12.2056366

“MOAO first on-sky demonstration with CANARY”, E. Gendron et al. A&A 529, L2

“Comparison between observation and simulation of sodium LGS return flux with a 20W CW laser on Tenerife”, R. Holzlöhner et al, Proc. SPIE 9909, Adaptive Optics Systems V, 99095E (2016); doi: 10.1117/12.2233072

"The ESO transportable LGS Unit for measurements of the LGS photon return flux and other experiments", Bonaccini Calia et al, Proc. SPIE 8450, (2012); doi:10.1117/12.926898,

“CANARY: the on-sky NGS/LGS MOAO demonstrator for EAGLE”, Richard M. Myers et al, Proc. SPIE 7015, Adaptive Optics Systems, 70150E (2008); doi:10.1117/12.789544

Image of the transportable ‘Wendelstein’ sodium laser guide star unit in its new enclosure just after sunset. The Nordic Optical Telescope can be seen in the background.