87th AMS Annual Meeting

Tuesday, 16 January 2007: 5:30 PM
Study of tropospheric turbulence using observation of shadow bands during a total solar eclipse
210A (Henry B. Gonzalez Convention Center)
Hari Om Vats, Physical Research Laboratory, Ahmedabad, Gujarat, India
The models predict that turbulence is initially generated on the largest scales troposphere and that the dissipative forces cause the turbulence to be degenerate into smaller and smaller scales, eventually being dissipated on scales much smaller than those with which we are concerned to the present case. The turbulence imposes random refractive index variations in the atmosphere which produce phase fluctuations (amplitude fluctuations are much less significant) in the wave front of the signals passing through the medium. In case of troposphere it is the optical signal coming from distant sources e.g. stars at night and very thin arc of the Sun just before and immediately after the total solar eclipse. The later is more relevant to the present case. The randomly phase modulated wave while propagating gradually manifest fluctuations in amplitude and intensity as well. These variations are seen by the observers in case of stars as twinkling of stars and in case of solar arc as shadow bands. Because the phase fluctuations are random and move with ambient medium twinkling as well as shadow bands vary both with time and space.

The nature provides us an exciting opportunity just before and after a total solar eclipse to see an interesting phenomenon of shadow bands. During many total solar eclipse of the past these have been seen. These bands can be easily distinguished by naked eye, which perceives moving patterns as enhanced against a stationary background. These bands are rather disorganized, but do tend to form a quasi linear pattern almost parallel to the tangent to the centre of solar crescent. This justifies their common name. In the early days observers have seen these and usually make an artistic picture based on the observations and some have made marks on pre-erected screen on the site of totality. There was non availability of really observable characteristics as temporal and spatial frequency distribution of the pattern, the persistence of individual features in the pattern and the speed of their motion, if any. The paucity of real scientific records of this phenomenon led to the proliferation of exotic proposals to explain the shadow bands. Now it is known that these patterns are linked to the phenomenon of stellar scintillation usually termed as twinkling of stars and the propagation of light through the tropospheric turbulence near the ground. Booker and Vats (1985) applied the refractive scintillation theory to the laser propagation through the atmosphere near ground level. This approach seems to be more applicable to the shadow band observations of total solar eclipse of 23 November 2003. The path of totality was over Antarctic region during which Sun's elevation was from grazing to a maximum of 15 deg, at the observing site during the observations it was ~ 2.7 deg. Thus it was like a horizontal propagation of light from visible crescent Sun before and after totality.

The shadow bands were clear seen for 4 minutes immediately before the totality and for 7 minutes after the totality. Thus this has been the longest duration of shadow bands ever observed and digitally recorded. Here we report the observation in the form of movie and present the analysis of the shadow bands of this eclipse. The cross correlation and spectrum analysis of the shadow band patterns observed and recorded near Maitri during the total solar eclipse on 23 November 2003 were performed. This eclipse took place, when the Sun was at grazing angle from the local horizon. The duration of shadow band patterns were very much extended before and after totality. The spectrum of the observed shadow band fluctuation indicate the presence of Kolomogrov or power law (with a spectral slope ~ - 2) type turbulence in the local troposphere. This matches fairly well with theoretical work, however, differs significantly from the spectra of shadow bands observed at different locations usually with higher elevation angle of the Sun. Codona (1986) defined two regimes for the temporal intensity spectrum slopes as -5/3 and -17/3. The present spectral shape is quite in contrast, possibly due to the observing geometry and the location. The preliminary correlation analysis gives a typical characteristic scale size of ~ 35 cm and pattern drift speed ~ 3.8 m per sec.

Supplementary URL: