The Mesosphere is interested by important chemical and dynamic phenomena, so that observation of its Sodium layer has became a common target of several research plans all over the world. In order to extend the study of its dynamical and chemical  variation we want to observe the Sodium emission during daytime and perform differential analysis looking for periodic changes. Using a small telescope, a Magneto-Optical Filter and a CCD camera onboard the International Space Station, we avoid the uncomfortable LIDAR technique taking advantage of the solar flux to stimulate the Sodium emission.











1.1  Introduction-Rationale
1.2  Research’s goals
1.3  The technique
1.4  Signal-to-Noise ratio
1.5  Data Analysis


2.1 The magneto optical filter
2.2  Image Sensor 


View/Download (2,412,651 bytes) Word zipped file

View/Download (8,109,557 bytes) PPT and AVI zipped files










Earth Observations


Observations of the Atmospheric Sodium Layer @ 90 Km in the Mesosphere




Using an imaging technique we want to extend the observation of the Mesospheric Sodium layer to many points of a large bidimensional region at the same time. So far, in order to partially achieve this goal, networks of LIDARs have been conceived from the ground, but our way is to go to space and use a very narrowband Sodium filter  to reject the rest of the unwanted solar spectrum. The solar light replaces the laser beam to excite  the Sodium atoms.





1.1 Introduction-Rationale


 The atmosphere of the Earth contains a layer of metallic Sodium. Even though almost 30 years have passed since the first routine investigations by  Gibson, Standford and Bowman  have been made [1 and 1bis], we still have no definite answers about its origin, large scale properties and temporal evolution. This lack of knowledge is the basic reason why several research groups, all around the world, are working on this field and, definitely, constitutes the main trigger for our proposal.Neutral atoms of Sodium have been first discovered by Slipher in 1929 [2] and they are located between 80 and 105 Km  in the Mesosphere (Fig.1 and Fig.2). The peak density of its sodium layer is around 90 Km at a level of 109÷1010 atoms/m3 and its investigation faces some difficulties:

  • its altitude is not suitable for balloons (too high) or for orbiting satellites (too low)


  • the background solar daylight radiation, diffused by the atmosphere is far exceeding its faint yellow glow, leading to  the problem of the signal to albedo noise, discussed in section 1.5

Fig. 1 Relative highs of various componrnts   



Fig. 2 Pictorial composition of the International Space Station with the Earth below, where the relative altitude and thickness of  Na layer and blue atmosphere at the limb are shown.


As far as the first point, LIDARs (systems composed by a powerful laser which excites the atmosphere, coupled with a fairly large telescope to collect the backscattered radiation from the various components) have made possible accurate observations, limited however to point-like targets only, so precluding large scale field of view (FOV) and spatio-temporal studies of its evolution.
As far as the second point, routine observations are usually being performed at night time only, so precluding long and uninterrupted temporal series to get better frequency spectra of its variations. Attempts have been made during the day, reducing the  divergence of the LIDAR beam or using narrowband Fabry-Perot filters that, however, are affected by extreme thermal and mechanical instability.

In spite of the above difficulties, the Mesospheric Sodium is the most frequently studied metallic component of the Earth's atmosphere because of its high backscattering cross section,  ( 7·10-17 m2ster-1, and abundance, 3·1013 atoms/m2 [2] ). However, so far, we are not able to respond with certainty to the basic question of  its origin, but can only list two possible groups of Sodium sources:

·                    Terrestrial Sources ( volcanic dusts and salt particles from the oceans)

·                    Extraterrestrial Sources (meteorites and comet dusts with an additional component of gas from the sun)

Most likely, ablation of meteorites and comet dusts is the main source mechanism; however, difficulties arise when trying to justify the observed abundance. Moreover, researchers find that the Sodium layer is changing as a function of geographical coordinates, seasons and even time during the day. Complex chemical reactions are  taking place that involve many ions and neutral particles at high levels: all depending from the basic parameter governing the evolution, that is the temperature (Fig. 3). 

Fig. 3 Temporal series of temperature profiles.


 Finally, one important aspect is its dynamical behaviour, driven by the tidal effects and gravity waves . These effects, with some contribution of atmospheric turbulence, are mainly responsible for short periods evolution. The available data [1 bis] show oscillations of the Sodium layer: in particular,  density and thickness  oscillate in phase, while the  altitude is 180 degrees out of phase.  Moreover, sometimes, sporadic layers are detected  with very high Sodium density and short formation time, whose origin is still not understood (Fig. 4).

Fig. 4 Density profiles.



1.2 Research’s goals


Because of its high resonance cross section, the Mesospheric Sodium constitutes a very good mark able to reveal waves and short duration transients phenomena in the Earth atmosphere. For this reason the main purpose of our experiment onboard of the ISS is to image and record such events in order to:

·                    further clarify the source of Sodium from meteorites and comet dusts detecting its consequent density variation . Since meteoric events are randomly localized in space and time, we would need a continuous monitor of a large FOV.

·                    get evidence, and possibly obtain spatio-temporal Fourier analysis, of waves, both tidal and gravity waves

·                    try to combine our space observations with other techniques from the ground. We plan to do this not only using data available  from  LIDAR routine measurements  but mainly from another particular project we are in collaboration with: this is the 'Earthshine Project '  conducted at the Big Bear Solar Observatory and Caltech in California. Form this project, it is possible to derive the absorption profile of terrestrial Sodium comparing  the illuminated side of the Moon with the dark side of it. Indeed, this last side can see the solar light after reflection by the Earth (Earth-shine). Since this light travels twice through the Earth's atmosphere, it is possible to derive the absorption effect of the Sodium layer  provided the observations are made at high spectral resolution. Fig. 5 shows example of such spectra kindly provided by the Earthshine Project. 

Figure 5a:   Figure 5b:

The two plots on top are the sodium D2 line spectra from the illuminated side of the Moon (left) and the dark side of it (right). The bottom plot shows their ratio. This last is affected by the terrestrial Sodium absorption. The spectral resolution, however, is coarse compared with the Sodium  line and the MOF bandwidth (about 40 milliAngstrom). The temperatures and columnar densities that is possible to derive are useful data  for other purposes also,  particularly  Adaptive Optics [5].

  variations with the atmospheric temperature of the scattering total cross section in the resonant transition D2 of the Sodium hyperfine structure.


1.3 The technique


Referring to Fig.6 where our instrument is shown as composed by three parts (a telescope, the MOF filter and an image sensor), we want to stress here the importance and the characteristics of the MOF (Magneto-Optical Filter). Indeed, the MOF is the core of the system that makes this experiment unique. Its working principle is described in section 2.1 and will be provided by our adviser, Prof. Alessandro Cacciani of the University “LA SAPIENZA” of  Rome. Here let be enough to say that it is a very stable and narrow band filter, about 50 milliAngstrom, achieving unsurpassed performances of perfect  tuning in the core of the Sodium yellow lines. Its weight ( about 1 Kg) and  dimensions (10x10x10 cm3)  are also attractive for space applications. As a filter, it will be located between the telescope and the image sensor (provided by JPL). In this manner we are able to reject all the other wavelengths of the solar spectrum so that we can definitely say that the MOF produces an artificial night, which is the necessary condition to detect the faint yellow glow originated by the Mesospheric Sodium. The most  important consequence of this artificial night is that we can work in the presence of daylight,  using  the solar radiation, instead of the uncomfortable LIDAR apparatus, to get the Sodium excited; moreover, the  solar radiation is not localized,  as the laser beam is, and we can observe many points at once of a large bidimensional FOV. A drawback of this otherwise ideal situation is that we must compete with the albedo radiation from the Earth inside the same wavelength pass-band of the MOF. This problem is discussed in the following section devoted to the Signal-to-Noise ratio.


Fig. 6  Mechanical assembly of the instrumentation


1.4  Signal-to-Noise ratio  


 Here we need to compute the expected signal and the competing luminosity from the Earth. 

As far as the expected signal, it comes from the Sodium atoms . We have:


The product of them gives the probability that one single solar photon is scattered back from the Sodium layer inside one steradiant. This amounts to 2·10-3 events per second and steradiant. 

We need to compare this number with a similar probability coming from the other components of the Earth, namely :  the clouds, the oceans, the lands and the atmosphere.  Each component displays a different albedo luminosity (clouds are white and oceans are blue), therefore we want to select the observations in the most favourable condition for the yellow part of the spectrum: that is, above the oceans and the air as a background. Unfortunately, albedo data in the narrow MOF band width are not available: only  integral or wide band spectral resolution data have been estimated in the whole solid angle (Fig. 7). 

It is, therefore, possible that  most of the measured yellow albedo from the oceans or the air is coming from the high resonance radiation of the Sodium atoms (which is our signal).

The above consideration, makes us aware that our proposed observations from the ISS are unique and  may be useful to interpret the albedo behaviour measured by the “Earthshine Project” (planned  collaboration).




  Fig. 7 Earth’s energy budget


To give some numbers, let us take the yellow albedo of  the air (with the exclusion of the Sodium resonance ) to be of the order of  10-2, that is less than 10-3 in a single steradiant.  This would  mean that the above quoted resonance probability is larger than the albedo in the same spectral bandwidth. We add that at the terminator (the twilight zone) the Sodium layer @ 90 Km  is excited by the solar radiation while the Earth’s atmosphere is not completely illuminated: so that the situation is even more favourable .

As a final step, in order to select a suitable detector,  we need to estimate the total number of photons reaching our image sensor . This can be done multiplying the resonance probability by the solar photon flux  inside the MOF bandwidth. It amounts to 7.4·1015 photons/s/m2  [7] which must be further scaled by the acceptance solid angle of our apparatus.  Considering an altitude of 400 Km for the Space Station and 90 Km for the Sodium layer, we get 10-13 steradiants for a 5 cm aperture telescope, and 4 times as much for 10 cm aperture, giving a useful  photon flux  of  about 3·103   from each square meter of the Sodium layer  seen from the Space Station. We need now to convert this number into pixel illumination of the image sensor.  The magnification factor of the telescope optics is given by the ratio D/F  (D=400 Km distance of the Space Station, F= focal length of the telescope) so that, considering  a typical figures of 10 microns for the pixel size and 0.4 meter for F, we find each pixel collecting photons from 100 m^2 of the Sodium layer.  This brings the photon flux to  be 3·105 photons /s per pixel, a fairly measurable quantity even  considering a further reduction of a factor 10 caused by the overall quantum efficiency of the Optics ( MOF included). Actually, the  photon flux must be increased by about a factor two to  include the albedo contribution. Note, moreover, that when we look at differential signals between frames, the albedo background is cancelled. 

The image sensor for this experiment will be provided by the Jet Propulsion Laboratory (JPL) of  Pasadena, California (sect. 2.2).



1.5    Data Analysis


The first order data analysis is to perform suitable time lag differences in order to get evidence of transient events, specially during the times of Leonids, Perseids etc. In this manner we can visualize the location and geometrical configuration of any glow increase. The temporal coincidence with the periods of known meteor swarms, will witness about their credibility.  Then we can try to evaluate the consequent density variation in the nearby  regions. Depending on the system sensitivity to small variations we can also try to get evidence of waves and look at  spatio-temporal  Power Spectra of the observed surface oscillations [6] (see also Fig. 8 as an example of monochromatic solar image through the MOF, showing  faint Sodium emissions otherwise invisible).

As a final remark, giving the perfect spectral stability of the MOF, we are able to scan across the Sodium emission  line, taking advantage of the spacecraft relative speed, ranging from
 -7 Km/s to + 7Km/s, well covering the whole line profile by Doppler shift.

This kind of  information, if prolonged in time,  are also useful also for other disciplines   connected with the very important problem of  global changes.

Fig. 8 An example of Sodium image obtained with the MOF: note the visibility of the Sodium emission regions. 






2.1 The magneto optical filter


The Magneto-Optical Filter (MOF) is a particular instrument that gives a really narrow bandwidth,  high transmission (almost 50%) and perfect stability. It can work only in a small range of  wavelengths, well defined, like Na and K doublets, and can be thought as an high resolution spectrograph within those wavelengths.

It was developed by  Alessandro Cacciani  at the end of the ’60s (see references and images in Its main use is for the study of the sun; during the last years it has been adopted also for the analysis of  Jupiter oscillations  and to observe the mesosphere in daylight (paper in preparation ).

A complete MOF instrumentation  is  made of two separated unit: the MOF itself and the WS (Wing Selector) that, together,  provide accurate Doppler and magnetic measurements. In this proposal,  however, we just need the first simple unit. It is composed by two crossed polarizers (P1, P2) and a metallic vapour (Sodium in our case) between them in a longitudinal magnetic field (2000¸4000 Gauss).  A schematic diagram of the filter is shown in Fig. 9.


        Fig. 9  Schematic diagram of the MOF filter (top) and spectral behaviour along the optical path. Following the Zeeman rules, the vapour immersed in a magnetic field absorbs two circularly polarized components, leaving the residual light also circularly polarized (in opposite direction). This wavelengths cannot be stopped by the  crossed polarizers and are transmitted as a couple narrow bands that can be close at will (depending on the magnetic field strength). 


The working principle of the MOF  is based on two concurrent effects, namely the Zeeman effect in absorption as discussed in Fig 9 and a sort of  Faraday rotation close to the line’s wings called Macaluso-Corbino effect. Both change the polarization in and around the resonance lines, leading to total transmission profile  typically  shown in Fig. 10.  

Fig. 10 experimental transmission’s profile at different temperature and magnetic field values: in this experiment we use a single band  transmission 10c. 



2.2  Image Sensor


The information in this  paragraph are from Dr. Cunnigham of JPL [7].

The Active Pixel Sensor (APS) is a CMOS-based visible image sensor

Because CMOS parts can be manufactured on demand and because CMOS is the process presently used to make almost all digital and most analog components, the APS can be fabricated as a "camera on a chip" where all of the timing and control circuitry are integrated on chip with the imaging circuitry. This allow the construction of a single chip digital camera.

For this effort the APS imager would be used behind the Sodium filter.

This would allow the APS to directly image the sodium line emitted from the atmosphere. It is envisioned that the APS would take exposures on the order of 1 second duration, in which approximately 1000 photons would be collected by each pixel. The quantum efficiency of scientific-quality photodiode active pixel sensors is approximately 50% and the noise floor

is approximately 50 electrons rms, which is more than adequate for this application. The APS would digitize the signal directly on-chip, and would transmit the image using a standard digital interface .





[1]     BOWMAN, M.R., GIBSON A. J., STANDFORD M. C. W., ’Atmospheric Sodium measured by a tuned laser radar’ Nature, 221, 457-459, 1969.

[1bis] CONTENTO, CARLA, “Studio preparatorio di strumentazione per misure di Sodio mesosferico in luce diurna”; Tesi di Laurea, Dipartimento di fisica, Università degli Studi di Roma “ La Sapienza”,  26/9/2002.

[2]     SLIPHER, V. M., “Emission in the spectrum of the light of the night sky”, Publ. Astron. Soc. Pacific, 41, 262, 1929.

[3]     HUMPHREYS, R. A., L. C.  BRADLEY, J. HERRMANN,  “Sodium-Layer Synthetic Beacons for Adaptive Optics” ; Lincoln Laboratory Journal Special Issue on Adaptive Optic, volume 5, number 1, 45, Spring 1992.

[4]     FOUKAL, P., “Solar Astrophysics”, John Wiley & Sons Editors, p. 67, 1990.

[5]     PATRIARCHI, P., A. CACCIANI,  "Laser Guide Star Adaptive Optics: Measuring the Sodium Density Using a Magneto-Optical Filter"  Astron.& Astrophys, 344, L45, 1999.

[6]     HARVEY, J., “Helioseismology: The State of the Art”, Fourth Soho Workshop ESA SP-376 April 1995

[7]      Dr. THOMAS J. CUNNINGHAM, private communication, 2002.

      Acting Group Supervisor

      Advanced CMOS Sensor Technology Group (387-I)

      NASA Jet Propulsion Laboratory

      Ph. (818) 354-0891

      Fax (818) 393-0045