GRAVITY Instrument Description

Overview

GRAVITY is not a monolithic instrument. It is a collection of sub-systems that aims to precisely control the wavefront of the incoming light and its path through the VLTI system before the actual combination of beams takes place and interference fringes are created.  A unique aspect of GRAVITY, and the first time this is ever realised, is its ability to interfere the light coming from either a single astronomical source (single-field on-axis) or from two sources simultaneously (dual-field). In dual-field, the GRAVITY system can perform phase referenced observations supported by the accurate knowledge of the path length which is assessed by a laser metrology system. In this mode of observation, the interferometric phase of the primary star is calibrated to that of the secondary against the detrimental influence of the atmosphere. This enables highly accurate angle measurements on sky and is the basis for GRAVITY's astrometric observing mode. In GRAVITY's imaging observing mode, dual-field observations allow to observe relatively faint targets and use a brighter star for fringe tracking. The difference between the two modes of GRAVITY is the observation strategy and the use of the laser metrology to measure the light-path.

The sub-systems of GRAVITY include:

  • The IR wavefront sensing system CIAO. CIAO is located in each of the UT coude room and it operates with the deformable mirror of the VLTI.
  • A polarisation control system to counteract polarisation effects in the VLTI. GRAVITY can work either in a split or a combined polarisation mode.
  • An active pupil guide system including LED sources mounted on each of the telescope spiders.
  • A field-guide system to track the position of the source and ensure proper injection into mono-mode fibers.

The overall GRAVITY system interacts and works in harmony with the VLT-I system of telescopes and delay lines. The operation of various GRAVITY sub-systems is transparent to the user. A detailed description of the working principle of GRAVITY with an example on how the system is arranged for on-sky observations can be found in Section 3 of the User Manual.

The single and dual field modes

The Beam Combiner InstrumentThe Beam Combining Instrument (BCI) with its sub-components labelled.

The primary unit of GRAVITY is the Beam Combining Instrument (BCI) that performs the acquisition process and provides the interferometric fringes. The GRAVITY BCI is cryogenically cooled and physically located in the VLT-I laboratory. Within the BCI cryostat, a field is separated and two stars find their way into either the science channel or the fringe-tracking channel. The GRAVITY fringe-tracker (FT) forms an integral part of the observational approach, i.e. GRAVITY science observations are always done with active fringe-tracking. The FT fringe position is analysed at a frequency of approximately a kHz in order to correct for the atmospheric and instrumental piston (i.e. a residual optical path difference between beams) by modulating piezo mounted mirrors within the instrument. The FT star thus allows longer detector integration times in the science channel (SC, up to 60 seconds) without compromising the contrast of the fringe pattern.
In single-field on-axis mode, the FT channel and SC channel receive light from the same star which is split 50%-50% by a beam splitter. In dual-field mode, the fringes of one star are formed in the science channel, and those of the other star in the fringe tracker channel. The two astronomical sources can have an angular separation up to 30''. For separations up to 0.6'' and 2.7'' on the UTs and ATs, respectively, the light can be split using the beam splitter so that only 50% of the light from each source are injected into the FT and SC channels ("dual-field on-axis"), respectively. For separations between 0.27" and 2.0" (UTs) or 1.17" and 4.0" (ATs), a roof-top mirror can be used and 100% of the light of the FT and SC objects reach the respective fibres ("dual-field off-axis"). For larger separations between the FT and SC objects of up to 30", the VLTI Star Separators are used to separate the beams for the two stars. This GRAVITY "dual-field wide" mode only provides relative measures (differential visibilities, and phases, and phase closures), no absolute visibilities.

GRAVITY delivers spectrally dispersed interference fringes that allow stellar interferometry. The FT spectrometer always operates at low spectral resolution (R ∼ 22). Taking advantage of the longer integration times, the science channel records the entire K-band at each of the three implemented spectral resolutions of R ∼ 22, 500 and 4000. The fringes give access to interferometric quantities such as absolute and differential visibility, spectral differential phase and closure phase. These quantities provide information of physical phenomena at a spatial resolution that can reach 2 mas (depending on the VLT-I baseline), as well as time-resolved differential astrometry at the exquisite accuracy of a few tens of μas.

Astrometry and phase-referenced imaging

Astrometry aims at measuring the separation between two targets, whereas phase-referenced imaging aims at measuring the phase of the SC target in reference to the FT target. Both techniques use the dual-field mode of GRAVITY and rely on the laser metrology to make the connection between the SC and FT measurements. For astrometry, the metrology measures the evolution of the differential optical path difference as a function of time, which through the interferometer baselines can be converted into a separation between the targets. For phase-referenced imaging the metrology is used to relocate the FT target reference to a separation offset close to the SC target, which helps for producing images. As the laser metrology provides relative optical path measurements only, a metrology zero must be determined for both techniques to work. This determination is typically achieved by swapping a pair of targets (i.e. reversing the sign of the separation), which separates the sidereal metrology signal that changes sign, from the constant metrology zero. When swapping is not directly possible on the target pair of interest, e.g. due to a very faint SC target not observable with the FT, the metrology zero determination can be carried out on a more balanced nearby pair. Other than the requirement to calibrate the metrology zero-points, astrometry and phase-referenced imaging observations are similar to dual-field observations and have the same observing constraints.

If you are interested in using the astrometric mode we invite you to have a look at the following articles: Gravity collaboration, 2017, A&A 602, A94 and The Messenger 170, 10. In addition, if you need help with the preparation of GRAVITY astrometric observations you can contact the ESO Helpdesk well in advance of the proposal deadline.

Acquisition templates from Period 112 on

In Period P112 the GRAVITY single-field and dual-field acquisition templates will both be split into two separate templates: on-axis and off-axis templates.

The single-field on-axis template has the same behaviour as the old single-field template. The new single-field off-axis template will be used when the user wants to observe a faint target and is only interested in using the fringe tracker to record data. In this mode all the light is sent to the fringe tracker and the science fiber is on sky.

Dual-field on-axis and off-axis templates are the same as the old dual-field template, except now the two modes are in separate templates and there is an overlap in the target separations that can be observed with these templates. Dual-field wide template remains the same. The possible target separations for the different dual-field modes are shown in the Figure below.  

The GRAVITY dual-field separationsPossible target separations for the different dual-field acquisition templates: on-axis, off-axis and wide.

More information on the new templates and separation limits can be found in The GRAVITY User Manual and The GRAVITY Template Manual.

Characteristics and performances  

The limiting magnitudes currently offered are listed in the two tables below. They cover four turbulence regimes, because the performance is largely influenced by the efficiency of the injection into the fibres. However, for certain targets it may be advisable to set more stringent, intermediate turbulence constraints, e.g. if offset Coudé guiding is required, or if the target is at high airmass (especially because the turbulence constraint is measured at zenith).

Table 1. K-band limiting correlated magnitudes on the ATs
    T ≤ 10%
seeing ≤ 0.6"
τ0 > 5.2ms
T ≤ 30%
seeing ≤ 0.8"
τ0 > 4.1ms
T ≤ 50%
seeing ≤ 1.0"
τ0 > 3.2ms
T ≤ 85%
seeing ≤ 1.4"
τ0 > 1.6ms
single-field on-axis SC (=FT) 9.0m 8.5m 8.0m 7.0m
single-field off-axis FT (no SC) 9.5m 9.0m 8.5m 7.5m

dual-field on-axis

(sep. < 2.7")

FT 9.0m 8.5m 8.0m 7.0m
SC 16.0m 15.5m 15.0m 14.0m

dual-field off-axis

(1.17'' < sep. < 4.0")

FT 9.5m 9.0m 8.5m 7.5m
SC 16.5m 16.0m 15.5m 14.5m

dual-field wide

(4.0''<sep.<30.0")

FT 9.5m 9.0m not offered
SC 14.0m 13.0m not offered
Table 2. K-band limiting correlated magnitudes on the UTs
    T ≤ 10%
seeing ≤ 0.6"
τ0 > 5.2ms
T ≤ 30%
seeing ≤ 0.8"
τ0 > 4.1ms
T ≤ 50%
seeing ≤ 1.0"
τ0 > 3.2ms
T ≤ 85%
seeing ≤ 1.4"
τ0 > 1.6ms
single-field on-axis SC (=FT) 10.0m 9.5m 9.0m 8.0m
single-field off-axis FT (no SC)
10.5m 10.0m 9.5m 8.5m

dual-field on-axis

(sep. < 0.6")

FT 10.0m 9.5m 9.0m 8.0m
SC 17.0m 16.5m 16.0m 15.0m

dual-field off-axis

(0.27'' < sep. < 2.0'')

FT 10.5m 10.0m 9.5m 8.5m
SC 17.5m 17.0m 16.5m 15.5m

dual-field wide

(2.0''< sep.< 30.0'')

FT 10.5m 10.0m not offered
SC 17.0m 16.0m not offered

Important note on the performance of the GRAVITY dual-field wide mode: The magnitude limits for GRAVITY dual-field wide are strongly dependent on the atmospheric conditions and separation between the science and fringe tracker target. E.g., for the UTs, stable fringe detection can be expected for a K=16mag science target 3 arcsec from the fringe tracker star and for T<10%, while no fringe detection can be expected for a K=14 mag science target 30 arcsec from the fringe tracker star and T ~ 30%. The given magnitude limits for the fringe tracking target are hard limits, whereas the sensitivity for the science targets are based on the preliminary data collected during commissioning runs and assume ~1 hour on target. Due to the need for good atmospheric conditions (turbulence class 30% and better), observations in service mode are strongly encouraged to optimize the scientific return.

Tables 1 and 2 above list the limiting K-band correlated magnitudes offered for observations with the ATs or UTs starting with P112. The following restrictions apply:

  • Low spectral resolution is offered in both dual and single field. However, users are advised not to use low spectral resolution in single-field on-axis, since the FT already provides low resolution data on the science target. Low resolution is now dual-field wide mode.
  • To be able to track all six baselines, the correlated magnitude has to fulfil the above criteria on at least three baselines that do not form a triangle (e.g. 13/23/24, or 12/13/14).  Additionally, in service mode the visibility should not drop below 5% for these three baselines and not below 1% in visitor mode.
  • Observations with GRAVITY should typically ask for THN sky conditions. For fainter targets CLR conditions should be requested.
  • For observations with GRAVITY no Moon (FLI) constraints ('n') should be requested. For FLI>0.85, a minimum moon distance of 10 deg (V<9) or 20 deg (V>9) should be used during observations, well within the conditions defined as bright time.
  • On ATs in dual-field off-axis mode, the separation of the two targets must be at less than 4", see discussion above for the exact separation limits for on-axis and off-axis modes.
  • On UTs in dual-field off-axis mode, the separation of the two targets must be at less than 2", see discussion above for the exact separation limits for on-axis and off-axis modes.
  • In dual-field wide mode the separation of the two targets must be less than 30".
  • In dual-field wide mode the science target has to have magnitude H<14mag on the ATs and H<17mag on the UTs to allow for the target to be centered using the acquisition camera.
  • While the limiting correlated magnitudes depend on the fringe tracker performance and are hence independent of the spectral resolution setting of the science channel, please consult the GRAVITY ETC to ensure sufficient SNR on the latter, especially in atmospheric absorption lines.
  • The science channel will reach saturation for the minimum DIT at a K-band magnitude of -1 for medium spectral resolution. At high spectral resolution and in split polarization the limiting brightness is K=-4. These brightness values apply for observations with the ATs.
  • For UTs, the estimated brightness limits are K=+4 and +1, respectively. In practice, the values depend on the strehl ratio achieved by GPAO or CIAO.
  • For VLTI specific limiting magnitudes for Coudé guiding, please refer to the latest version of the VLTI user manual.

More detailed information on the VLTI facility is available in the latest version of the VLTI user manual. More information on the quality control performed during the night and also on the calibrated data can be found from the quality control pages.

High Contrast at Close Seperations

GRAVITY can provide high contrast observations at close seperations using the dual-field capabilities. High precision astrometry and direct spectoscopy are possible for a variety of science cases. A description of how to acheive the best possible results can be found in Pourre, N. et al. (2024).

GRAVITY Contrast Curves
Plot from Pourre, N. et al (2024) Detection limits of some representative instruments dedicated to direct observations of exoplanets. GRAVITY detection limit corresponds to the inner part of the on-axis mode of the instrument only. It is limited to 27 min exposure time and companions oriented parallel to the longest baselines of the VLTI. Plot adapted from Dr. Bailey script available at https: //github.com/nasavbailey/DI-flux-ratio-plot. The HST/STIS curve is from Ren et al. (2017). VLT SPHERE/IFS IRDIS is from Langlois et al. (2021). VLT SPHERE/SAM is from Stolker et al. (2023). VLT JWST/NIRCam is from Carter et al. (2023). (Blue triangles) Estimated reflected flux in the visible for exoplanets observed using the radial velocity technique. The estimation follows a Lambertian model with radii fixed at 1 RJup and a geometric albedo of 0.5. (Red triangle) Estimated infrared flux for mature exoplanets observed using the radial velocity technique. Computed from equilibrium temperature estimates and planet radii fixed at 1 RJup. All visible and infrared estimated fluxes are based on Traub & Oppenheimer (2010).