Laser guide stars are artificial beacons created by focusing a laser into the upper atmosphere. They are used to power adaptive optics systems on large telescopes to image objects in space. There are two types of laser guide stars, Rayleigh laser guide stars, which use Rayleigh backscattering from air molecules in the upper atmosphere and Sodium laser guide stars, which use resonant backscattering from sodium atoms in the mesosphere.
[Return to top]  [Return to home page]

TABLE OF CONTENTS    palomar laser launch telescope

Who needs Laser Guide Stars?
Cone Effect
Rayleigh Laser Guide Stars:
            Rayleigh Guide Stars on Big Telescopes
Sodium Laser Guide Stars:
           Where does the Sodium come from?
              Interaction between the Sodium Atom and the Laser Beam
Lasers for Sodium Guide Stars
CW Dye Laser
Sum Frequency Laser:
           Pulsed Mode-Locked Sum-Frequency Laser
              Measured photon returns as a function of date
              CW Mode-Locked Sum-Frequency Laser
              CW Single Frequency Sum-Frequency laser
              Fiber Lasers
Increasing the photon return/watt
             Chirping Experiment at Palomar Observatory
             Backpumping Experiment at Palomar Observatory

Who needs Laser Guide Stars?

Adaptive optics uses a distance source outside the earth's atmosphere to measure its wavefront distortion.  However, because atmosphere is three dimensional, the wavefront distortion introduced by this turbulence is different in different directions. This effect is called anisoplanatism and limits the field of view of an AO system. For astronomical sources the field of view is very small ( less than a milliradian) . A star bright enough to drive the AO system has to be with this field of view of the target source, greatly limiting the utility of AO for defense and astronomical applications. The solution is to generate a artificial laser guide star using a laser pointing in the direction of the target. Laser guide stars are used by astronomers to observe faint sources and by the military to image or illuminate objects in space.
[Return to top]  [Return to home page]

Cone EffectCone effect

All laser guide stars suffer from another type of anisoplanatism called the Cone effect (or focal anisoplanatism). This is  shown schematically in the figure on the right.
Because the laser guide star is at a finite height, light from the Laser Guide Star travels through a different path to light from the science target and so measures a different wavefront error, integrated through the atmosphere. The figure shows that the effect scales as the  telescope size and the ratio of the height of the turbulence to the laser guide star. As a rule of thumb, we obtain good correction if the path difference between light from the space target and the laser guide star as less than 1 radian at the height of the turbulence. The size of telescope that can be corrected is given by

                                                                do equation

High altitude turbulence exists at 10 to 15 km, so this equation gives us an approximate size of the telescope that can be corrected with a single beacon. Large astronomical and DOD telescopes require high altitude beacons.

[Return to top]  [Return to home page]

Rayleigh Laser Guide Stars

Rayleigh laser guide star at MMTRayleigh laser guide stars use backscattering of the laser from air molecules in the upper atmosphere. The fraction of light backscattered depends on the fourth power of the frequency and typical Rayleigh laser guide star systems either use a frequency doubled Nd:YAG laser (0.532 microns) or an Eximer UV laser. The brightness of the Rayleigh laser guide star falls rapidly with height, setting an upper limit of about 25 km to the beacon height.  Rayleigh laser guide stars are being used by Steward Observatory [see photo left]. The laser generates a short pulse and the detector is range gated to observe only the focused image, which is a few arcseconds in diameter. This technology will be used by the CAMERA system, under development at Palomar Observatory, for a 1.5 meter diameter telescope.
[Return to top]  [Return to home page]

Rayleigh Laser Guides on Big Telescopes

Rayleigh guide stars are highly suited for correction of atmospheric turbulence near the ground ( aka "Ground layer Adaptive Optics")for large telescopes, since they do not measure high altitude turbulence, which is thus uncorrected. For some sites, such as in Antarctica, most of the turbulence is near the ground and improved, though not diffraction limited, imaging performance can be obtained over a much wider field of view.
[Return to top]  [Return to home page]

Sodium Laser Guide Stars

Much higher altitude beacons can be obtained by focusing a laser, accurately tuned to the sodium D2 line, into the mesosphere, a layer of the atmosphere about 90 km above the ground. Free sodium metal exists in this region and can be used to form a bright beacon by resonant scattering of the laser light. Sodium atoms produce the yellow light of street lights. In the street light the sodium atoms are excited by the  electric current passing thought the tube.  For a laser guide star the atom is produced by absorbing light from the laser beam and remitting the light at a similar frequency. There are two D lines available for generating the sodium guide star, the D2 line has twice the cross-section to the radiation field and has always been used for this application.
[Return to top]  [Return to home page]

Where does the Sodium come from?

  About 100 tons of meteorite burn up in the upper atmosphere every day, evaporating about 100 kg of sodium metal into the upper atmosphere. The total mass of sodium available in the entire mesosphere for generating a guide star is about 500 kg, implying a lifetime of about 5 days. There is significant annual and diurnal variation in the sodium density and enhanced short-lived sodium abundances, so-called “sporadic” sodium, occur on shorter time scales. Such enhancements can increase the abundance by over a factor of ten for periods of minutes to hours. Typical column densities are between 2 and 7 x10^13  atoms/m^2 with peak densities sometimes exceeded 20 x 10^13 atoms/m^2. It is remarkable that so little sodium metal can backscatter even a few per cent of the radiation from a laser beam.
[Return to top]  [Return to home page]

Interaction between the Sodium Atom and the Laser Beam

The spectral and temporal characteristics of the laser must be carefully tailored to the sodium atom if we are to obtain a reasonable photon return from the mesosphere. The relevant physics of the sodium atom for laser guide stars is discussed on more detail at this website on a separate page. The trick is to excite the sodium atoms so that the time to absorb and re emit a photon is about 100 nsec when the atom and laser frequency are well matched. Different laser technologies of the same average power can produce photon returns that differ by more than a factor of ten, so understanding the physics is as  important as understanding how to  build a laser. In summary there are two competing effects:

Saturation  Because the sodium atom spends a finite amount of light in its excited state there is a limit to the number of photons that can be produced by a sodium atom however powerful a laser is used to excite it. This results in saturation of the photon return and becomes an important effect for a single frequency laser at an intensity of above 50 watts/m^2.

Optical Pumping. Light changes the electronic state of the atoms, giving rise the various optical pumping effects. Some effects increase the return, others pump the atoms into a state where they can no longer interact with the light. The polarization of the light, strength and direction of the earth's magnetic field all affect the degree and type of optical pumping. The photon return of broad band CW lasers  is more adversely effected by optical pumping than a single frequency CW laser. Pulsed broad band lasers are less effected by optical pumping because the atoms have time to achieve thermal equilibrium between pulses.
[Return to top]  [Return to home page]

Lasers for Sodium Laser Guide Stars

CW Dye Lasers

Nestor Farmiga with CW dye laser at Yerkes The first Laser Guide Stars produced with a commercial CW laser were made in Feb 1992  by my group using the finder telescope of the 40 inch refractor at Yerkeslaser guide stars Observatory to launch the laser. The laser is shown on the left with an undergraduate, Nestor Farmiga. The laser was a Spectra-Physics CW dye laser pumped with a second hand 20 watt Argon-Ion laser.  The laser generating about a watt of power and were able to obtain a laser beacon about 2 arcseconds diameter in the sky. You can see the beam leaving the finder of the 40 inch refractor at Yerkes Observatory. The image of the guide star, taken through the 40 inch refractor itself (with additional displacement of the laser), is shown on the far right. The return was a bright as a 12th magnitude star in the V band. In this image, the "finger of God" is due partly to Rayleigh backscattering from the lower atmosphere but the red blob ( a false color image) was due to the volcanic eruption of Mount Pinatubo in June 1991. The ash layer was then situated at a height of 23 km.

Chicago dye laser at MMT 1993The same laser was taken to the original MMT and fired close to the secondary mirror (see left). This is change in launch position important for large telescopes since this position produces the smallest spot across the telescope pupil. Much of the low altitude backscattering is obscured from the AO system by the secondary mirror. Rayleigh backscattering is always a problem with CW systems and becomes an even more serious problem when multiple beacons are used, due to an effect called fratricide. The return was not sufficiently high to use for our AO system and we switched to the sum frequency laser described in the next section. Although the CW dye laser is not now the technology of choice,  ESO have developed an operation laser, PARSEC, (2007) on the Yepun 8 meter  VLT telescope. This laser currently generates over 12 watts and uses a fiber feed to connect the laser to the launch telescope.
[Return to top]  [Return to home page]

Pulsed Sum Frequency Lasers

The dye laser is the only readily available laser that can be tuned to the sodium D2 line. Most of the current generation of lasers used in astronomy employ two infra-red Neodymium YAG lasers, one operated at 1.064 micron and the other at 1.319 micron that are combined in a non-linear crystal to form yellow light. Sum frequency generation of lightThis technology was first developed by Tom Jeys in MIT/Lincoln Labs.This was a long pulse mode locked laser with an almost ideal spectral and temporal format, providing a high photon return from the sodium layer with the ability to gate out low altitude Rayleigh backscattering and backscattering for Cirrus clouds. Tom Jeys built a very successful 2.5 watt 10 Hz flash-pumped laser that obtained the first clear observations of optical pumping, in collaboration with our group. He built a diode pumped version of the laser for us in 1994 which was tested on the Dunn Telescope at Sacramento Peak MN on 10 nights in Dec 1997 through Jan 1998. The average return was 740,000 photons/sec/m^2/Watt in an average column density of 7x10^14 sodium atoms/m^2.
[Return to top]  [Return to home page]

Chicago/Palomar Pulsed Sum Frequency laser Laser gain engine

We have improved the original Tom Jeys design into a reliable laser that has been used at the 200 inch telescope at Palomar Observatory for the last four years. The most difficult technology was the laser gain engines. We use a zig-zag slab clamped between two sapphire plates that are conduction cooled via water cooled copper heat sinks. An early version, illustrating the basic design, is shown right. The Nd:YAG slab is pumped by two sets of Coherent QCW diode lasers mounted by LLNL. The slabs are side pumped in a double pass configuration. The current laser technology generates 10 watts of TEM00 light at 1.319 micron  and 16 watts of 1.064 micron. Both lasers are mode locked, generating 1 nsec pulses @ 100 MHz under a 150 microsecond pulse envelop. The pulse frequency is typically 400 Hz.
Sum frequency converter
Each IR laser produces  nanosecond pulses with a peak power greater than 1 kW. At this power level, sum frequency conversion is an efficient process even with bulk materials such as LBO. We use three LBO crystals in series, with reimaging between crystals to produce 8 watts of yellow light.
Laser mode locked pulsesThe pulse formats are shown right. The top trace is the diode laser pump pulse, next is the 1.06 micron laser pulse, then 1.319 micron and finally the yellow light pulse. The pulse structure shown is an artifact  cased by aliasing between laser pulse  and oscilloscope sample frequency. The initial spike is due to relaxation oscillations caused by the pulse nature of the pump. These oscillations are damped out using an intra cavity non-linear crystal.

This laser technology is simple and robust. It was built at the Observatory site. It has no active controls save for the temperature of the LBO non-crystals and operates in an dusty environment with  temperature changes of a 10 degree being common over an observing period. It has proved to be highly reliable in service.
[Return to top]  [Return to home page]

Measured Photon Returns as a Function of Date
Photon return/watt as function of date
Measured photon returns as a function of date are given left. The vertical axis is the number of photons/cm^2 received at the telescope/ watt of laser power delivered to the sodium layer. They do not included the effect of atmospheric absorption on the return signal (typically 85% transmission). Most of the observations were carried out in summer months, when the sodium column density is low. The wide variability of the sodium abundance is one reason why comparisons of photon return/watt between different lasers has proved so difficult.
[Return to top]  [Return to home page]

Photon returns for Different Types of Laser

A chart giving measured returns of existing lasers is given on the Center for Adaptive Optics laser website. This website also contains links to other laser guide star groups. We can make some observations from this data. Firstly CW single frequency and long pulse lasers have a higher return than CW broad-band lasers. The reasons why are explained in the tutorial of the Physics of the Sodium Atom. In summary, a key parameter in efficient excitation of the sodium atom is its cycle time, or the time it takes to absorb and re emit a photon. For a single frequency CW laser,for a cycle time of 100 nanoseconds we need an laser intensity of about 40 watts/m^2, corresponding to a laser power of about 10 watts for typical guide star. If the light from  CW laser is divided amongst a number of lines, the cycle time becomes long enough that the Earth's magnetic field undoes much of the beneficial effect of optical pumping. Various other effects also tend to depopulate the number of sodium atoms available to interact with the laser beam. This is the reason that a single frequency CW laser has a much higher photon return/watt than a broad-band or multiline laser.

 In contrast, the Pulsed multiline laser used at Palomar has a much higher intensity during the pulses, and so can efficiently optically pump the atoms. Furthermore, the time between pulses ( 2.5 milliseconds) enables the sodium atoms to rethermalize before the next pulse and repopulates the upper ground level. Detailed calculations by Milonni show that the photon return/watt should be similar to a single frequency CW laser. These calculations are confirmed by direct measurements and my own calculations.
The SOR laser has operated at a site with relatively poor seeing, generating a fairly big spot at the sodium layer. The higher powers, or smaller spots, saturation becomes a more serious problem, although it occurs at a higher level then was generally believed. The effect of saturation is discussed in more detail at another page. In summary, although the center of the line starts to saturate round about 40 Watts/m^2 ( 50% return at an intensity of 63  Watts/m^2), the linewidth broadens at higher powers, which means more sodium atoms interact with the laser beam. This reduces the effect of saturation ( see link for more details)
[Return to top]  [Return to home page]

CW Sum Frequency Lasers

Lockheed Martin Coherent Technologies

  A  12 watt CW mode locked sum frequency with a spectral bandwidth of 550 MHz, built by Lockheed Martin Coherent Technologies has been operational at the Gemini North Observatory for over three years.A 40 watt version with a 1.7 GHz bandwidth has been delivered to Keck observatory and a 50 watt version will be shipped to Gemini South Observatory. High  efficiency in the sum frequency conversion is obtained by using short mode locked pulses ( .3 nanosecond pulses @ 12 nsec repetition rate). This however increases the number of lines generated by the laser and this is expected to  significantly reduce the photon return/watt.
[Return to top]  [Return to home page]

 CW Single Frequency Sum Frequency Laser

The most successful CW single frequency laser currently operation is the 50 watt laser at the Starfire Optical Range. This uses two cw single frequency lasers that are fed into a ring cavity containing the LBO sum frequency crystal. This cavity builds up very high power levels and results in high conversion efficiency. The single frequency CW operational results in a high photon return. This laser is being developed commercially by FASORTRONICS.
[Return to top]  [Return to home page]

Fiber Lasers

Fiber lasers represent the newest technology, using CW fiber lasers and Raman amplifiers to generate high power. ESO have demonstrated a 25 watt version with higher powers been recently reported.
[Return to top]  [Return to home page]

Increasing the photon return of laser guide stars

Understanding of the basic physics of the interaction between the sodium atom, its environment and the laser radiation field allows the laser designer to improve the photon return/watt from the laser guide star. The Physics of the sodium atom is outlined in another page. Two techniques for improving the photon return have been proposed by MIT/Lincoln Labs. The first is called chirping, which involves changing the laser frequency as a function of time, the second is called backpumping, which uses a second laser frequency to pump atoms from the lower round state to the upper ground state. We have recently carried out experiments at Palomar Observatory to test these ideas in July and October 2009. The results are in fair agreement with theory and predict that we can build lasers it reasonable cost with significantly higher returns/watt than the current generation of lasers.
[Return to top]  [Return to home page]

Chirping and Backpumping Experiments at Palomar Observatory

The Pulsed Sum-frequency laser at Palomar Observatory was modified to carry out chirping and backpumping experiments in 2009. The Optical layout of the laser is shown below.

Optical layout of sum frequency laser
[Return to top]  [Return to home page]

Chirping Experiment

Chirping was achieved by mounting one of the fold mirrors in the 1.06 micron laser cavity onto a PZT mount. The mount was driven by a sinusoidal voltage, phase locked to the laser pulse frequency. The relative phase shift between the laser pulse and sine wave was adjusted so that the mirror crossed its zero position at the time of the laser pulse. It was possible to change the phase of the drive signal by 180 degrees so that the laser could be either red or blue shifted with time. The position of the mirror was calibrated using a Michelson interferometer mounted into the beam. This allows us to adjust  the chirp rate to a predetermines number of MHz/microsecond.
Results of chirping Experiment
The results of the experiment carried out in July 2009 are shown right. The photon return was first measured with no chirp, then with a preset chirp, and then with no chirp. The no chirp signal was obtained from the average of the two no chirp measurements. The chirping enhancement defined as the ratio between the photon return with chirp divided by the average no chirp photon return. Measurements were made redshifting and blue shifting the laser frequency with time. Also shown is the predicted enhancement obtained from the Monte Carlo Code.
This experiment showed an impressive enhancement of 1.8 at a chirp rate of 0.6 MHz/microsecond. When blue shifted light was used there is no, or perhaps some small negative enhancement, as would be expected theoretically.
[Return to top]  [Return to home page]

Back-Scattering Experiment

Backscattering experiment return v frequency

In a second experiment, he output of the 1.319 micron laser was also fed through a resonant cavity EOLM tuned to 1.71 GHz. This produced two sidebands, one of which was able to pump the sodium atoms from the F=1 ground state into the F=2 ground stateThese observations were made under poor seeing conditions ( 2 arcsecond). We were however able to get enhancements ( defined in the same way as the chirping experiments) of a factor of 1.2 centered on a frequency of 1.718 GHz with3% sideband power. There was no enhancement at 1.71 GHz and the resonant nature of the EOLM meant that we could not push the tuning beyond 1.722 GHZ.

backpump enhancement v power

Having found the correct tuning frequency we measured the enhancement as a function of power in the side-band. This is shown in the figure right. These measurements are consistent with theory if we assume that atmospheric transmission was low (60%) and the spot size was 2 arcseconds.Photon return experiments were carried out at Palomar in October 2009, at the time of intense forest fires under poor seeing conditions and limited the time available.
[Return to top]  [Return to home page]

7:08 pm Feb 23 2010 Edward Kibblewhite