THE ESO AO SYSTEM: NAOS-CONICA (NACO)

NACO is the VLTs Main Adaptive optics system (located on the 4th 8m telescope (Yepun or UT4)). Consisting of a 196 element Shack-Hartmann visible (0.5-1 micron WFS) AO system (NAOS; the blue ring below) and a 1-5um diffraction-limited camera (CONICA; the red dewar below).

SIMULTANEOUS DIFFERENTIAL IMAGING: Planet Hunting with NACO-SDI

In a collaboration between Laird Close (Steward Observatory) & Rainer Lenzen (MPIA) and ESO has led to the development of the NACO SDI device. NACO-SDI solves the contrast problems with direct extra-solar planet detection. If one could image a planet in and out of the deep 1.62 um methane absorption feature then one could subtract the images and reveal a methane dominated planetary atmosphere while removing the contaminating light of star (see Close et al. 2004; Lenzen et al. 2004 proc. "ESO's workshop on science with AO" , Garching, Sept 2003).

Below is our discovery image of the closest brown dwarf binary Eps Indi BaBb (Blue=1.2um, Green =1.6um, Red=2.1um).

As NACO H spectra of the newly discovered T-dwarf  binary Eps Indi Ba/b show T-dwarfs (Teff<1200 K) have significant methane absorption (This data was taken during first light of NACO SDI just after we discovered this exciting brown dwarf binary). The cooler the object the stronger the stronger the absorption (Ba a T1 is 1240K, 43 Mjup; Bb a T6 is 850K, 27 Mjup; McCaughrean et al. 2003 astroph0309256). A young extrasolar planet is predicted to have similar spectral features to a T-dwarf.

A cartoon of how you could place the 3 filters across the Methane feature.
In  this manner we should have excellent contrast with respect to the primary star

 
 
    To best  correct for the chromatic effects of the speckle noise it is required to use 4 different filters all at the same time. Laird Close has designed an optical element (a double wollaston) which can produce 4 identical images of the same star at the same time (see image below).

As well we observed Titan the methane rich moon of Saturn.

See Hartung et al. A&A 2004 for our SDI paper on Titan.
See here for our Titan press releases
See here for our MOVIE of titan from the SDI data

Here we see Rainer Lenzen (MPIA), Max (ESO), and Markus Hartung (ESO) installing SDI  into CONICA.

Here is the new f/40 SDI camera (Note the custom focal plane "quad" filter; designed by Laird Close) inside the dewar (f/40 camera designed by Rainer Lenzen; PI of CONCIA).

After comm. 1 we successfully installed these devices into NACO. The following data were obtained with NACO.
 

A key to having a very good final subtraction is to not have any "non-common path" aberrations between the 4 SDI beams. Markus Hartung (ESO) has shown with phase diversity that the non-common path errors are less than 10nm between the 4 images (this is very close to perfect).

Even after 2 minutes of observing there is still a large amount of speckles (super speckles NEVER average out -- you are always speckle-noise limited). However, the SDI device samples and subtracts this "speckle pattern" very well in each of the 4 beams.

Compared to just added the images together (Classical AO) the SDI device can subtract out the speckle noise and hence is ~10-100 times higher contrast than normal AO. The data reduction is done by a complex script produced by graduate student Beth Biller (Steward Observatory).

Above we show the line of fake planets (10,000 times fainter T6 planets)
blinking from 0 to 33 degrees, along with the derotator.

Our initial commissioning data from the VLT (as reduced by graduate student Beth Biller) suggests that we are close to photon noise limited from 0.5" outwards. This is over an order of magnitude better than was possible before. Hence we can now detect (at 6 sigma) a planet ~20,000 times (delta H=11)  fainter than its star at separations of only 0.6" in just 40 min of telescope time. These are the highest contrast astronomical images ever made (from the ground or space).

Based on these photon-noise limited contrasts (detecting a companion ) we predict the following sensitivities for a 2 hour exposure on the VLT