ALFA laser system

Adaptive optics with a Laser For Astronomy

ALFA is an adaptive optics system with a laser guide star, built at the Calar Alto 3.5-m telescope as a joint project between MPE and MPIA. This page provides a details about the design and operation of the laser, and ALFA's involvement in the European laser guide star network.

Although the Adaptive Optics module for ALFA is still being used for observations at Calar Alto using natural guide stars, the laser has been decomissioned. We are now working on a more powerful, remotely operable version called PARSEC which will be used at the VLT.

	ALFA laser system
Results latest: PARSEC (Oct 00) previous 23% Strehl (Jun 99) NGC 6764 (May 99) UGC 1347 (Aug 98)		system design beam diagnostics observing with a LGS movies of the LGS TMR involvement (LIDAR experiment)

What you see here is 4 W of laser power, enough to produce an artifical guide star in the sodium layer 90 km up that can be used to greatly improve the quality of images by correcting for turbulence in the atmosphere. Low down, the beam appears bright due to the backscatter from gas and dust.

For a list of full results, publications, team members, how to apply for observing time, etc. see the ALFA homepage

system design

The beam is produced by a continuous wave ring dye laser with a typical output of 4 W, which is pumped by a 25 W Ar⁺ laser. The dye laser is tuned to the Na D₂ 589.2 nm line and has a bandwidth of only 10 MHz. Before it leaves the coudé lab, the beam is circularly polarised and pre-expanded to about 10 mm diameter. It is then directed along a beam relay to the launch telescope mounted beside the main telescope. Here the beam is magnified to about 15 cm across and projected into the atmosphere so that it can be focussed on the Na layer. These stages are shown schematically on a photo which highlights the main features of the system. The laser and beam relay are briefly summarised in an on-line overview which explains the choice of lasers, where they are installed, and how they work.

Such a powerful laser pointed into the air can, of course, be dangerous for aircraft pilots if they happen accidently to look down the beam. Although they wouldn't be permanently blinded, they would certainly be dazzled. ALIENS is an aircraft detection system, which has been installed to avoid this danger. It is simply an optical camera mounted on the front ring of the telescope, which shuts off the laser when there is a detection of a moving object.

beam diagnostics

A diagnostics bench has been built directly below the launch telescope in order to monitor the quality of the projected beam. This has proved essential during optimisation of the LGS. It measures a number of important parameters:
1) Beam position and angle: to insure that the beam is projected into the launch telescope on axis.
2) Beam jitter: has shown that the jitter the laser acquires as it passes through the beam relay is negligeable.
3) Beam transverse profile: to check that the beam still has a Gaussian profile, which is needed to give the smallest possible spot size on the WFS.
4) Total power: to monitor the transmission of the beam relay (up to 75%), and give some idea of how bright the LGS should appear on the WFS.
5) Beam collimation & static abberations: measured by a Shearing interferometer to make sure that the beam reaching the launch telescope has a high quality, and show whether any optical elements are of poor quality and need replacing.
6) Dynamic wavefront errors: measured by a Shack-Hartmann sensor (CCD and lenslet array) to look for beam degradation caused by, for example, temperature differences in the coudé lab and dome.
7) Polarisation: measured at the top of the launch telescope to keep the out-going veam circularly polarised since this produced a brighter LGS.
8) Starcam: to look at stars through the launch telescope so that it can be kept perfectly aligned.

observing with a LGS

A single-window GUI, the Laserleitstand, controls the essential aspects of the entire laser, beam relay, diagnostics, and launch telescope. The Laserleitstand has been designed so that using the laser is relatively straight forward. However, life is not so simple when things go wrong, and so there is a user's guide (6.3M postscript or 2.2M gzipped) to take you step by step through the system: trouble-shooting, optimising the laser, etc.

Before acquiring the laser on the WFS it is important that the telescope is focussed as well as possible, because the height of the sodium layer (and therefore also the focus position of the WFS) is not known exactly. This step is not necessary when using a natural start since the WFS focus position is always the same; the perfect telescope focus is where the WFS residual is zero.
A set of automated algorithms have made the task of acquiring the laser on the wavefront sensor very simple. First the laser is viewed on the TV-guider, which produces an excellent picture of the Rayleigh cone tipped by the LGS (it also acts as a very sensitive `cloud detector' and any thin layer of moisture in the atmosphere shows up clearly as a bright ring in the Rayleigh cone). Once the LGS is positioned correctly, the TV-guider is then moved out so the LGS can be seen on the WFS. The positioning can then be fine-tuned. Both the launch telescope and WFS focus are adjusted roughly by hand, and then a pair of automated algorithms obtain the perfect focus positions so that the spots are as small and bright as possible. A reference fibre is then moved into the laser focus position (another automated algorithm finds this position since it is not known in advance - it depends on the distance to the sodium layer) in the AO bench and used to calibrate modes on the WFS. Once calibration is complete, the fibre is moved out and the system is ready to be used.

The LGS brightness is around 9 mag, although it is rather dependent on zenith angle and Na column density. That is enough for the adaptive optics to make good improvements in image quality.

movies of the LGS

Until mid 1999 the LGS always appeared very extended on the wavefront sensor, as can be seen in a WFS image, which provide a comparison of the laser and a natural guide star. Movies from March 1998 of the natural guide star (NGS) and laser guide star (LGS) show the open-loop motion of the spots, using 2 lenslet arrays (3x3 & 5x5) and sampling frequencies (60 Hz & 100 Hz):
MPEG format: NGS 3x3 60Hz LGS 3x3 60Hz NGS 5x5 100Hz LGS 5x5 100Hz

A more recent LGS 5x5 array movie from June 99 shows the enormous improvement in the appearance of the LGS. The spots are now much more compact and brighter, resulting in significantly better centroid measurements. To see just how much improvement has been made, it is worth seeing the image taken when we were able to first lock on the LGS in Sept 97. The spot in the centre shows that the system wasn't aligned properly then.

The jitter of the spot pattern is completely due to the atmosphere, the combination of tip-tilt from different upward and downward paths. It can best be seen if the LGS is imaged through a single lens, as this LGS single lenslet movie from May 1999 shows. This is a problem becuase the LGS can move too close to the edge, and even out, of the centroiding regions causing the AO loop to crash.

involvement in the TMR network

ALFA in the TMR

While it was running, MPE was a partner of the European `Training & Mobility of Researchers' network on Laser Guide Stars for 8-metre class Telescopes, which coordinates the main drive of the European effort in this area. ALFA was involved in a number of experiments pertinent to the use of lasers for adaptive optics, including tilt recovery from a LGS, sodium layer statistics, atmospheric turbulence, light pollution, etc. These experiments and some of the preliminary results are described in more detail on the ALFA TMR page.

The structure and full work of the network is summarised in a paper by Foy, the coordinator of the network: ``The Laser Guide Star TMR network of the European Union'' (200kB gzipped postscript), given at the ESO conference in Sept 1998 devoted to astronomy with adaptive optics.