The Eclipse Expedition to Jalu/Libya 2006 Mar 29

by Dr. Wolfgang Strickling

 Diese Seite auf Deutsch

Fotos  |  Shadow bands  |  Weather  |  Technic  | Weblinks

This eclipse was excellent! We have had extreme air transparency, nearly no wind and a position almost exact on the centerline. During totality, i got the impression, that the eclipse is very bright. The reason might be the bright desert sand, reflecting light from the borders of totaliy into the umbral zone. Unlike my last eclipse in Zimbabwe it was possible to read camera displays etc. without a torch.

Other observes outside the desert, eg. in Turkey, did not confirm this observation. Venus was very bright and we could recognize it almost 15 minutes before 2nd contact. Mercury was hardly to see, Mars, Orion and other stars were not visible, although i did not seek intensively for them.

Shadow bands were contrastful and good to be seen. However the are hardly to be seen on my video, perhaps because of their higher movement speed, compared with my recordings of 2001. The approaching and vanishing of the moon's shadow was impressive.

The accompanying sunrises and sunsets were great! Some of our group members saw the green flash and the zodiacal light for their first time.

Our coordinates (position of my tent) were 28° 14.1123' north (=28.235205°), 21° 30.3337' east (=21,505562°), 138 m elevation, mean of 5 measurements, datum WGS 84. This was 1.3 km south of the centerline.

Contact times (local time = UT + 2h), calculated with AstroWin  ( for DeltaT = 64.9 sec) EmapWin and EclipseCalculator:

           h  m  s  Pa Nord  Pa Zenit  Azim  Elev  Sh.Band     EmapWin   EclipseCalculator 4:03.6     
1. Cont.: 11:08:31    225°    264°     135°   57°             11:08:36       11:08:24.8
2. Cont.: 12:26:38     46°     53°     173°   65°    138°     12:26:45       12:26:33.9
3. Cont.: 12:30:45    223°    228°     175°   65°    135°     12:30:48       12:30:37.5
4. Cont.: 13:50:05     45°     12°     218°   60°             13:50:08       13:49:58.9

Left: Baily's Beads at 2nd contact.

Right: Detail.

Left: Baily's Beads at 3rd contact.

Right: Detail.

Photos with shorter expore times show the inner corona and the prominences (left 1/250 s), longer exposed photos reveal the outer coronal structures (right: 1/20 s). The right image was processed with a radial unsharp mask.

All these coral images were taken on Fujicolor Senia 100 ASA. The advantage of a color negative film, compared to slide film, is its huge dynamic recordable range. I am surprised about the amount of details contained in even one picture.

My optic were mirror telephoto lenses (MTO 11 CA with 1000 mm f/10 and 500 mm f/8) on a Vixen NP equatorial mount.

Exposures with 4 or 6 seconds time show the lunar surface in the Earth's light.

Left: 4 s with 1000 mm f/10

The right image was exposed 6 s with my 500 mm f/8.

Combining several pictures with digital image processing software show the inner corona as well as it outer structures like streamers. I used 'Fitswork' for adding the photos and generating a Larson-Sekanina mask. Photoshop was used for fine processing. Click on the image to get a high resolution version.

With a video recording we made a polarization experiment. We placed a linear polarization filter in front of a video camcorder and rotated it. For the image i took frames from the video and assigned the vertical polarisation (position angle of electric vecor is 0°) to the red colour channel, position angle 60° to green and position angle 120° to blue. You see a strong radial polarization of the coronal light.

Many thanks to Reinhard Braun for his video equipment!

A polarization photo of the midlle corona shows Fred Bruenjes. His image has to be rotated appr. 45 ° clockwise to have the same orientation as mine.  Very detailed is a study of the coronal polarization by the Gundogdu-Observatory , Turkey.

Shadow Bands

The shadow bands were very impressive to observe. I prepared a cloth with 1,45 x 2,33 m extension, aligned to 317° azimut, the expected orientation of the shadow bands. I filmed the cloth with a mini-DV camcorder (Sony PC 100E) for about 15 Minutes around totality. On the video, the shadow bands are not very good to be seen.
Left: unprocessed image from my video, 25 sec before 2nd contact.

For image processing details see my shadow bands page!

Below: Images after video and image processing.

The wavelength of the shadow bands sunk from 45 cm in the beginning (2 minutes befor 2n contact) downto 6-7 cm a few seconds before 2nd contact. Finally the shadow bands were very fine structured. For the image right i superimposed a grid of 50 cm extension. There was a high turbulence in the shadow bands and they had a short survival time of 0.2 s or less.

After 3rd contact the shadow bands distance increased, like a reversal of the situation before 2nd contact. The graph left shows the wavelength (i. e. the bands distance) of the most  obvious bands in relation to the contact time.

The shadow bands moved from southwest to northeast, according to wind direction. There seemed to exist several components with different speed and wavelength. The main bands moved with 2-3 m/s, the fastest and largest components reached 6 m/s velocity.  This was similar to the measured wind speed at our location.

A video from 2nd contact 29.03.2006, 25 seconds on youtube.
Contrast enhanced video of the shadow bands before second contact.

To study contrast an the development of the shadow bands in detail, i took with  LIMOVIE the intensity profils of a four pixel area of my cloth. With  SPECTROGRAM i made a power spectrum of the intensity plot. The upper graphs show the spectrum befor resecond contact (marked as C2), the graph right the development after third contact (C3). It can be seen, that the shadow bands do not develop continuously, but that there may occur short periods of less activity. Longer measurements reveal, that first shadow band activity begins appr 3.5 - 3 minutes befor second contact.

A 10-seconds video MPEG2 file is here for download (sofi2006_shb2.mpg, 487 kB from 12:26:05 - 12:26:15 (UT + 2h)). The first second has a grid of 50 x 50 cm added as size reference.
(If you get problems in playing MPEG-2 files, download VLC media player or  Microsoft's newest newest media player. The standard one often does not play MPEG-2.)

Weather reordings:

With a Psion Organiser II as a data logger i measured the temperature, brightness and wind speed.

The temperature sunk from 31.3°C (at 11:13 local time) to 21.5°C (12:36 2 m above ground). Near the ground (0.1 m obove ground) it sunk from 34.5°C (at 11:13) to 21.5°C at 12:36. Unfortunately my data logger stopped 2 minutes before totality, so my sky brightness sensor did not take brightness probes during totality to have an objective measurement of the eclipse brightness.
However the shadow bands video and its exposure log recordings give a value of about 3 to 10 lux during totality, which was a bit brighter than the value i recorded in Zimbabwe 2001.

The left graphics show the wheather development on eclipse day, with temperature in 2 m and 0.1 m height, brightness and wind speed. There was only few wind, an eclipse wind at totality was neither detectable nor subjectively to feel.

My complete data tables are for download as CSV-file.

More wheather observations are published by Andrew J White. He also measured humidity. It decreased from 28% at first contact to 24% at fourth contact.

An idea of the good transparency gives the left picture of the green flash, observed at sunset of 2006 march 28, the evening before the eclipse. The green colour comes from the different refraction of light rays in our earth's atmosphere like in an optical prism.

Material and method:

As mentioned, a Psion organiser II recorded wheather measurements and triggered my photo cameras.
More details on my eclipse Organiser webpage (German) and my technical documentation of the interface (English).

my combined wind and temperature and brightness sensor.

1) Brightness

The sensor is the circuit "TSL230" (Texas Instruments), which converts light intensitiy to a frequency. It is mounted under a diffusor dome. The circuit can be varied in its sensitivity by changing the sensor chip area by a factor of 10 or 100. Additionally, the output frequency can be divided by 2, 10, or 100. So you get by simple adjustmend settings a dynamic range of more than 10^4. The output frequency can vary from < 1 Hz up to > 1 MHz. So it is possible to use one sensor for measuring light both for day and night. As the output is TTL-level, you can connet the circuit directly to a PC's parallel (printer) port or the input ports of my psion organiser extension box. The measured frequencies are computed and stored in a standardized manner due to sensitivity and output division settings.

2) Temperature

Temperature reagistration was done by the circuit "LM 75" (National Semiconductors) for ground registration. Its resolution is 0.5 ° C. The more precise DS 1621 provided data at 2 m height.

The temperature can be read via a two wire line interface (I2C bus) directly and digitally by the PC or organiser.

3) Wind

The anemometer had four hemisperic shovels.