Dr. Dale Simonich

INPE, Brazil
E-mail: Esta dirección electrónica esta protegida contra spam bots. Necesita activar JavaScript para visualizarla

 

The mesosphere is the region of the atmosphere where the temperature is lowest (the place where the minimum temperature of the atmosphere occurs is called the mesopause) and is located in the region of 80 to 100km. The region contains layers of free metals such as sodium, potassium, iron, calcium and others. The atomic transitions of these metals can be measured with resonance lidar techniques.

The mesospheric sodium layer has been observed with lidar since 1969. The reason for observing this metal is the very large resonant scattering coefficient of free sodium. It is possible to make useful measurements with lasers emitting tens of millijoules if the bandwidth is narrow and the receiver telescope is of reasonable area.

The mesosphere is a very active region both chemically and dynamically. Sodium (although a minor constituent of the mesosphere) is an integral part of these processes. Thus it is a very good tracer of these activities.

A sodium lidar to just measure the sodium density should have a power aperture product of at least about 0.005 Wm2 , a transmitted bandwidth of the order of 10 pm or less and a small receiver bandwidth (interference filters with a bandwidth of 5 nm are reasonable priced). The receiver must have photon counting capability. In order to improve the measurement of stratospheric aerosols a second low sensitivity channel is desirable because of the large dynamic range of the scattered signal. In order to do the final tuning (and also as a real time diagnostic and calibration tool) a sodium vapor scattering cell is needed.

Data reduction for the sodium density follows the standard procedure of using the lidar equation and calibrating the returned signal with reference to the Rayleigh signal at a height in the stratosphere where there is no aerosol. Since this is essentially a comparison of the Rayleigh backscatter crossection to the sodium backscatter crossection care must be taken when determining the latter. The Rayleigh return comes from the entire bandwidth of the transmitted signal while the sodium return only comes from that portion of the transmitted bandwidth that is within the sodium resonance line (hence the usual name of resonance for this type of lidar) of about 2 pm. In order to determine the transmitted bandwidth, and therefor the effective backscatter coefficient, one must use spectroscopic methods with a resolution of 1 pm or better on a portion of the laser output. If the laser is stable one needs to perform this calibration only infrequently.

Internal gravity waves generated in the lower atmosphere and solar thermal tides generated in the stratosphere are two types of atmospheric waves that propagate with upward group velocity components and growing amplitudes (because of the vertically decreasing atmospheric density) and are easily detected in the sodium vertical profiles. Because of the increasing amplitude of these waves they can break in this region and transfer their momentum to the atmosphere. Thus we have a means of transferring momentum from the lower atmosphere to the mesosphere. These studies of the momentum flux for the mesosphere are important in understanding the momentum balance of the whole atmosphere.

Modern interference filters have very low losses (essentially what ever is not transmitted is reflected). Thus one can use the sodium filter as a dichroic filter reflecting most of the pure rotational Raman spectrum into another receiver channel. Since Raman scattering only comes from atmospheric molecules and not from aerosols, one can obtain the atmospheric density trough the region of aerosol scattering if one has enough power. This allows one to have a true measure of the aerosol scattering and to do Rayleigh temperature measurements on the atmosphere through the region of stratospheric aerosols.

Because the sodium D2a emission is generated from quantum mechanical transitions, the backscattered photons are Doppler broadened both by the temperature of the atoms and their linear movement. Since the atoms are in thermal equilibrium with the atmosphere it is possible to measure the atmospheric temperature using them and also the radial component of the local wind transporting them. For temperature, the technique is to measure the ratio between the backscatter signals at two wavelengths. The wavelengths usually chosen are the peak of the D2a line and the minimum between the D2a and D2b lines. Measuring this ratio, which changes rapidly with temperature, makes it possible to determine the temperature. Winds are determined from measurements made at two points equidistant from the first wavelength. The temperatures obtained are very useful to atmospheric physicists because of their high vertical resolution. Lidar wind measurements also have better time and height resolution than other techniques.

The data analysis for temperature and winds is pretty much the same as that for density. The major differences are that the bandwidth of the emitted laser line is very much narrower, on the order of 0.1 pm, and the precision with which one must know the absolute wavelength of the emitted laser line must be on the order of 0.01 pm. The wavelength is usually measured using Doppler free spectrometry. The laser is scanned in small steps through the region of the peak of the D2a line and stopped at the dip in the Doppler free spectrum that corresponds to the strongest of the hyperfine lines. This gives an absolute wavelength reference for the first part of the measurement. The second wavelength can be obtained using an acousto-optical wavelength shifter to shift the above line to (or tuning the laser to) a point between the hyperfine groups called the cross-over point which again has a very precise quantum-mechanically determined wavelength.

In São José dos Campos we have used a less precise (but much cheaper technique) called a comb filter output coupler. A Fabry-Perot filter is used as the output coupler of the laser. Its free spectral range is exactly equal to the wavelength separation of the D2a and D2b lines. We first tune the Fabry-Perot to the peaks of the D2 line and make a measurement. We then tune to half way between them and measure again (any additional lines for this position are outside of the sodium resonance). We can then use the ratio of these measurements to obtain the temperature. The technique does not lend itself to measuring winds.

Temperature measurements in São José dos Campos, Brazil sometimes show a double minimum in the temperature. This double mesopause structure has been seen at other sites.

Two topics of current interest in mesospheric sodium are sporadic Na layers and curving striations in sodium. The first is a phenomenon that consists generally of very narrow layers of sodium that have peak densities many times those of the background sodium layer. These layers generally occur on the topside of the layer and are correlated with ionospheric sporadic E. The source of these thin layers is still not fully understood (they were first observed by the INPE lidar in 1977) although there has been some success with the theory of layering of sodium ions by wind shear and their subsequent neutralization. The structures generally descend with time. The time scale is very variable from a few minutes to many hours and the spatial extent varies from tens to thousands of km. The second phenomenon has only been noticed recently, being first reported in a paper on data from Puerto Rico in 1998-99 published in 2001. The phenomenon consists of highly sloping, curving striations in sodium with peak densities not very different from the background on the topside of the layer. They sometimes appear to have a wave like structure in time. One possible explanation offered is that breaking internal gravity waves produces them.