Chapter 5

Chapter 5, Technical Studies

Section 5.3: Sodium Layer Testing

NOAO Logo    Gemini Logo


Multi-conjugate adaptive optics (MCAO) is a technology that enables a GSMT to realize high spatial resolution, one of its hallmarks, over an extended field of view. Estimates of MCAO performance suggest that diffraction-limited point spread functions (PSFs) can be achieved at wavelengths from 1 micron longward over fields of view of at least 2 arcmin, and at Strehl ratios in excess of 0.2. This requires high-order correction of wavefront phase errors at high temporal frequency. The most direct path to enabling this level of correction requires laser beacons to provide multiple, bright, calibrating point-like sources.

A detailed discussion of the requirements for the adaptive optics (AO) system is found in Section 4.6.2. In this section, we note the parameters drawn from that discussion, which set the level of illumination required for adequate spatial and temporal sampling of the wavefront in order to achieve that level of correction.

Spatial sampling (sub-aperture spacing)0.5 m
Temporal sampling (WFS sample rate)800 Hz
Photon flux per guide star160 photons/cm2/sec

Taken together, these imply a guide star magnitude of order 10, clearly too bright for natural stars to be an alternative for multiple guide stars wherever we want to observe. Both Rayleigh beacons (produced by Rayleigh scattering of light in the atmosphere) and sodium beacons (produced by resonance fluorescence emission of laser light from the mesospheric sodium layer in the atmosphere) may play a role. (See Table 3, Section 4.6.2 for a discussion of the pros and cons of each.) The sodium beacon has one significant advantage over the Rayleigh beacon with regards to what is called the cone effect: it produces a guide star at a height of about 90 km, over four times the height of practical Rayleigh beacons. The conical beam returning from the sodium laser beacon overlaps over ¾ of the volume of the beam from an astronomical object at the same position in the sky. The conical beam from a Rayleigh beacon, coming from much lower in the atmosphere (less than 20 km altitude), samples a much smaller fraction of the atmospheric volume. Thus, for reasons of geometry (other things being equal), the sodium beacon is much more efficient in sampling atmospheric turbulence than is the Rayleigh beacon.

Figure 1 Laser light showing Rayleigh scattered 'plume' and guide 'star' and trailed background stars. The false color shows relative intensities. This image was taken from a 10-inch telescope adjacent to the laser room.

Clearly, sodium beacons are a promising technology for GSMTs. For this reason, we have been contributing to studies of the sodium layer in Chile, initiated by Gemini Observatory, with the specific purpose of preparing for the Gemini South Adaptive Optics project.

Whatever the details of the sodium laser selected, the number of photons returned per watt of power launched is the most important variable. This is directly proportional to the column density of sodium atoms in the mesospheric sodium layer. Various groups interested in atmospheric physics have studied this layer.1,2,3 The mesospheric physics is not completely understood, but the sodium atoms are presumed to be deposited there in a quasi-continuous process involving the breakup of meteorites. As such, there are questions about variability in the column density with (1) latitude, (2) season, and (3) time of night. These are important to the design of a sodium layer system because they set the minimum level of power required.

Another important parameter is the height of the layer. The sodium layer is not confocal with the science targets, and provision for making this differential focus adjustment must be built into the wavefront sensor (WFS) optics.

The sodium layer height is in fact observed to vary by ±10 km.1,4 The focus adjustments of the WFS must track this variation and take into account the variations in distance to the sodium beacon caused by change in elevation. The time scale on which significant height variation will take place will affect the design of the WFS focus system. In particular, it is important that the sodium layer height variations be slow and smooth enough to use the WFS to detect focus aberrations in the atmosphere.

Figure 2 300-mw laser launched from laser room on Cerro Tololo.

Finally, the sodium layer is observed to vary in structure and thickness, and to occasionally exhibit a multi-modal density in height.5,4 The most dramatic events are the sudden appearance of strong sublayers, called sporadics, whose origins have not been explained with certainty. This kind of structure will degrade the performance of the WFS, so it is important to know how frequent and severe these effects are. Also, because of the finite aperture of the GSMT, the elongation of the centrally launched beacon as seen from edges of the pupil is significant. The elongation of a 5-km-thick layer is 1.9 arcsec, substantially larger than the image of the seeing-broadened laser beacon. Although these effects have been observed before, they are likely to be at least somewhat dependent on latitude. Therefore, it is important to monitor the character of the sodium layer at the site where it will be used.


Gemini and the New Initiatives Office (NIO) have funded an effort to carry out five 7-10 day runs during the course of a calendar year at Cerro Tololo, close to the site of Gemini South where the MCAO system will be installed. Céline D'Orgeville of Gemini is the principal investigator. At this writing, three of the five observation periods have been successfully completed. The last run takes place in February 2002.

The program is being conducted in collaboration with the Imperial College (London) group, who has supplied a low-power (400-mw) dye laser with locking to the atomic sodium D2 line frequency, and a launch telescope to project a ~ 35-cm diameter beam vertically to the sodium layer. The laser and telescope have been mounted on a large optical table in a room on the summit of Cerro Tololo (15 km from the Gemini South site) and projected through a hole in the roof. During the first run, the escape of warm air caused some wavefront degradation in the launched beam. In subsequent runs, the lens in the launch telescope was sealed to the hole to prevent the warm updraft.

Figure 3 Sodium layer laser 'plume' viewed from the 0.9-m telescope. The plume is elongated because of the oblique view from the telescope. The intensity variations reflect the varying sodium density through the ~ 10-km-layer thickness.

The resonant emission at 589 nm is observed from two telescopes also on the Tololo summit: the 0.9-m telescope with 0.396 arcsec pixels, and the Schmidt telescope of 60-cm aperture and 2.3 arcsec pixels. The 0.9-m, located 140 m from the laser launch site, gives well-sampled images of the guide star profile, elongated by the ~ 10-km thickness of the layer and separation of the telescope from the laser. The Schmidt, 110 m from the laser and 41 m from the 0.9-m, is used to monitor atmospheric transparency and to provide the capability of taking images of the guide star simultaneously with the 0.9-m. A synchronizing signal is sent between telescopes that assures simultaneity of the exposures at the millisecond level. The shift of the laser plume with respect to the background stars as seen in the two images enables precise triangulation of the absolute sodium layer height. Meanwhile, the sequence of 0.9-m images monitors variations in sodium layer height, sodium layer thickness and structure, and-most importantly-total photon return. Launched laser power is monitored with a calibrated power meter at the launch telescope.

Reduction of the data is not complete, but the data appear of good quality, and there have been many hours on sky for each of the three runs so far. The data set will be a unique resource for the planning of laser guide star systems for the southern hemisphere, incorporating good seasonal coverage, long time base, and high temporal and spatial resolution. Figure 4 shows a sample of the data from the most recent run.

Figure 4. A sample of data. The 'Original Profiles' segment is a montage of a sequence of intensity profiles. The profiles are extracted from the raw images by summing the intensity across the width of the elongated laser 'plume.' The data are subsequently corrected for variations in laser intensity. The relative height variations are derived from the displacement of the plume relative to the CCD field. The data represents 500 frames of 10-sec integration, taken over ~ 4.5 hours.


First, it is worth repeating that this effort was an initiative of the Gemini Observatory. It has received considerable support from NOAO through CTIO, as well as from the co-investigators on the original proposal. The laser launch facility was developed by the Adaptive Optics Group at Imperial College (London), and the laser hardware was provided under contract by Imperial College to Gemini for the duration of the measurements from February 2001 to February 2002. In addition, we benefited from the experience gained by the Imperial College teams in their previous campaigns, and they were important in setting up the hardware and participating in the first run.


  1. Papen, G. C.; Garnder, C. S.; Yu, J. "Characterization of the Mesospheric Sodicum Layer". OSA Technical Digest Series 13, 96 (1996).

  2. Ge, J.; Jacobsen, B. P.; Angel, J. R. P.; McGuire, P. C.; Roberts, T.; McLeod, B. A.; Lloyd-Hart, M. "Simultaneous measurements of sodium column density and laser guide star brightness". Proc. SPIE 3353, 242 (1998)

  3. Clemesha, B. R.; Simonich, D. M.; Batista, P. P. "A long-term trend in the height of the atmospheric sodium layer - Possible evidence for global change". Geophysical Research Letters 19, 457 (1992)

  4. Michaille, L.; Cañas, A. D.; Dainty, J. C.; Maxwell, J.; Gregory, T.; Quartel, J. C.; Reavell, F. C.; Wilson, R. W.; Wooder, N. J. "A laser beacon for monitoring the mesospheric sodium layer at La Palma". MNRAS 318, 139 (2000)

  5. Ageorges, N.; Hubin, N.; Redfern, R. M. "Atmospheric Sodium Column Density Monitoring". Proc. ESO 56, 3 (1999)

October 2002