Wide spread cloud cover in many areas of Australia prevented planned observations of the partial lunar eclipse of 1995 April 15. However, successful video and visual observations were made from Bethanga, Victoria for 36 minutes and for the whole eclipse from Lavington, NSW and Ellenbrook, WA. Other observation reports included lunar feature timings from Victoria and WA. Analysis of the crater timings and of four video records to find the geometry of the umbra are presented.
Keywords: Lunar eclipse, crater timing, umbral geometry
The weather for the 1995 April 15 partial lunar eclipse was the worst experienced in the east of Australia for many years and was particularly disappointing for the last umbral lunar eclipse visible in Australasia until 1999 July 28.
However, a team of observers in Bethanga, a country town in the north of Victoria observed the eclipse during 36 minutes of crystal clear skies until they also experienced cloud cover. Another astronomer in the team observed at Lavington, NSW and recorded two video tapes of the complete eclipse with only short periods of cloud.
Dr Peter Skilton, observed his first successful lunar eclipse from his Lakewood Observatory near Frankston, Victoria and provided an interesting summary of his twelve lunar feature timings.
Brian Sture and Harry Moller also observed the complete eclipse from Ellenbrook near Perth in Western Australia, and provided an excellent video record for analysis, and Graham Wolf timed penumbral and umbral contacts from Wainuiomata, near Wellington, New Zealand, in clear weather conditions.
Maurice Clark observed from the grounds of the Murdoch University, in Perth, WA, while hoping to time a grazing occultation. Once again, Maurice exceeded expectations by providing 68 timings under "the most difficult conditions I have encountered", adding that "because it was a partial eclipse it made it very difficult to tell which way the shadow edge was going".
Peter Crowe reported an interesting 'naked-eye' observation in perfect skies from Anderson Miller's magnificent east-facing site at Gravelly Ridge in a remote Jarrah Forrest in Western Australia.
Other prompt reports were received of heavy cloud cover from most observers in Canberra, Sydney, the Blue Mountains, Frankston and Melbourne.
The unprecedented number of video recordings provided for analysis, together with the crater timings have made it a very interesting eclipse, as described in this report.
An Observers' Guide was issued in The Lunar Eclipse Observer, Volume 1 Number 4 (Soulsby, 1994a) well before the eclipse in 1994 October, which included a list of computer generated timing predictions for craters expected to be immersed in the umbral shadow, as well as some information concerning video recording techniques.
In response to a request from the Bethanga team organizer, Robert Price, a limited edition of The Lunar Eclipse Observer, Volume 1 Number 5 (Soulsby, 1995a) was issued to him in 1995 April, just one week prior to the eclipse. This special edition included detailed observing techniques and practical tips for the video recording of the lunar eclipse with some guidance on image analysis requirements.
A video-8 recording produced by Lindsay Millar and Robert T. Price of Bethanga (at longitude 147d 05' 22.8", latitude -36d 07' 10.9", site elevation 350 m) was received by the author for evaluation less than one week after the eclipse. They used a Sony HandyCam video camera, coupled to a Celestron C 11 SC electrically driven telescope on a German (equatorial) mount. The focal length of the C 11 SC was 2800 mm, and with a 30 mm eyepiece gave video images at a magnification near 93X .
John Bennett observed from Lavington, NSW, (at longitude 146d 56' 38", latitude -36d 03' 26", site elevation 180 m) and provided two separate video records a week later. For the first, John used a Celestron C 8 + with a CV 200 C surveillance video camera at f/10 prime focus to produce a VHS record, which was somewhat affected by internal reflections from the C 8. For his second record, John used a 700 mm focal length, 60 mm refracting telescope with a JS 12 S surveillance video camera recording to a Video-8 Sony HandyCam. His refractor was so sensitive it was masked down to 6 mm aperture, giving a somewhat 'soft' but acceptable focus. Both of John's video records were only partially affected by cloud and he captured the complete eclipse and included an audio timing signal on both video tapes starting at 12h 00m UT.
Brian Sture and Harry Moller recorded the complete eclipse from 11h 38m UT to 13h 02m with Brian's 100 mm f/6.3 Meade Schmidt Cassegrain telescope fitted with a number 12 yellow filter and a 400 line (horizontal) monochrome surveillance video camera feeding to a 250 line (horizontal) video recorder. The timing signal was an audio track from VNG and WWVH, and Harry repeated the time and announced drive adjustments when small tracking errors accumulated. Their observing site was near Perth at Ellenbrook (at latitude -31d 45' 50", longitude 116d 00' 33", elevation 50 m) with good weather, average seeing, a temperature of 17 to 12 C and a relative humidity of 80 %.
Valuable primary and feature contact timings were provided by Robert Price observing in Bethanga and Peter Skilton at his Lakewood Observatory (at latitude -38d 09' 49.3", longitude 145d 09' 28.2", elevation 87 m). Peter used a 15 cm f/5 equatorial Newtonian of 75 cm focal length and observed at 101X and at 29X for the emersion phase. Further timings were received from Graham Wolf in New Zealand who used a 60 mm terrestrial zoom refractor at 50X, and a record number of timings were received from Maurice Clark who observed in Perth with an 200 mm f/6 Newtonian using a 12 mm Nagler eyepiece.
Results of the analysis of crater and primary contact times from the author's 16 bit QuickBasic software and Macintosh IIci computer are given in Table 1. The values obtained for % enlargement and umbral oblateness derived, have been used to update the author's improved lunar eclipse ephemerides (Soulsby, 1990) for more exact predictions and for use in Cannons of Eclipses.
|
Observer |
Number |
%E |
Mean error |
Number |
%E |
Mean error |
|---|---|---|---|---|---|---|
|
Robert Price |
3 |
2.053 |
0.067 |
1 |
2.765 |
none |
|
Peter Skilton |
9 |
1.597 |
0.203 |
1 |
3.368 |
none |
|
Graham Wolf |
1 |
1.570 |
none |
1 |
2.930 |
none |
|
Maurice L. Clark |
23 |
2.036 |
0.082 |
39 |
2.212 |
0.058 |
Monochrome (Grey Scale 8-bit) and colour images (RGB 24-bit) were captured from three of the four video records with QuickImage version 2.2 software, a 24-bit frame grabber fitted to a Macintosh IIci computer for detailed image analysis.
One of the Bethanga images is shown in Figure 1, with other images in this series analysed by the examination of contour traces of the intensity, or lunar brightness of each, and the influence on this by the intrusion of the umbral shadow after first primary contact, as can be seen in Figures 2 and 3.
Crater timings are estimated at the most dense edge of the umbra where the change in shadow density is judged to be at a maximum. For this eclipse, where the umbra grazed the Moon, considerable importance was placed on accurate timing of the first and fourth primary contacts to provide the best data to determine the umbral size and its oblateness.
The primary contact time was determined by detailed examination of the limb darkening of the Bethanga series of 58 closely-spaced images (capture time was around 3 to 5 seconds each, a minimum for the QuickImage 24 software) and by other image analysis of the Lavington 60 mm refractor and Ellenbrook image sequences, using the same QuickImage software, frame grabber and Macintosh computer.
The Lavington and Ellenbrook images were compared using a system look-up-table from a standard RGB image. The false colour images were closely examined until the luna limb was just obscured by the umbra for first primary contact (1 C) and when the limb could just be seen for fourth primary contact (4 C).
The intrusion of the umbra was earlier than the general prediction, and gave a time of first contact at 11h 40m 26s from the captured Lavington images shown in Figure 4. The time of 1 C was predicted at 11h 40m 18s by the author's improved lunar eclipse ephemerides (ILEE) algorithm, described in Appendix C, 8 seconds earlier than the image time.
However, the end of umbral intrusion was 2m 23s later than the ILEE predicted time as shown in Figure 5 where a definite 4 C occurred in the false colour Ellenbrook series at 12h 54m, and in one false colour image in the Lavington series between cloudy conditions, also at 12h 54m.
These timings are used to further refine the umbral oblateness value, particularly as this eclipse was an extreme case, similar to the grazing partial eclipse of 1991 December 21, where change in oblateness is much easier to detect as the measured values are near the polar regions of the Earths' atmosphere where flattening is greatest.
The predicted first contact time was 11h 40m 30s using the classical oblateness value of 1/298.3, but the image analysis times for 1 C and 4 C indicate an increase in umbral oblateness. As the limb of the Moon at each primary contact is at the extreme southern edge of the umbra (near the north pole of the Moon - see Appendix A, Fig A-1), the effect of increased
umbral oblateness can be more noticeable than when the Moon traverses the equatorial region of the umbra. Due to increased oblateness and the reduce umbral polar semi-diameter, it was expected that 1 C time would be slightly earlier and 4 C time later.
The author has developed an algorithm (described in Appendix C) which enables the primary contacts to be calculated using the observed value of umbral oblateness, the motion of the Moon and its Besselian coordinates. With an input oblateness of 1/80, first primary contact for 1995 April 15 was expected to be at 11h 40m 18s and at 12h 56m 23s for fourth contact. The contact times found for 1 C from the video images shown in Figure 4 was only 8 seconds later than the time predicted by the AILEEP Apl computer program, but the time of 4 C, taken from Figure 5, was 2m 23s later than the estimated primary contact time.
However, the observations reported for the first primary contact were 11h 40m 35s (Price) and 11h 43m 37s (Skilton), and 12h 56m 47s (Price) and 12h 58m 03s (Skilton) for fourth contact. Both primary contact timings reported by Robert Price were very close to the contact predictions derived from the author's improved lunar eclipse ephemerides.
It was found that the Bethanga colour images were difficult to analyse by the technique of contour traces (Soulsby, 1993). However, one unsuccessful attempt to determine the umbral geometry using Image Analyst Revision 8.0 by best fit circles to the limb and umbra edge is illustrated in Figure 6.
With a calibrated lunar limb semi-diameter of 0.2733 degrees the field measured 0.143 degrees, which was only 26 % of the Moon's diameter. The gradient strength used for the umbral edge determination by the sign of first gradient, was a minimum of 1, to allow the system to find less defined edges and a maximum of 100 to eliminate any noise in the form of severe gradients. However, the limited field and insufficient image contrast at the umbral edge, made determination of the umbral size impossible.
When applying the Image Analyst technique to a Lavington refractor image a good fit to the umbra was found as shown in Figure 7. The umbral semi-diameter was measured as 0.413 degrees, which was typically quite low when compared with the theoretical value of 0.7532 degrees at mid-eclipse (Soulsby, 1994a).
The RGB Lavington image captured at 12h 30m UT shown in Figure 8, measured 0.546 degrees for the umbral semi-diameter, and when corrected for refraction and topocentric declination of the image as described in Appendix B, resulted in a value of 0.620 degrees.
Seven lunar eclipses have now been observed and recorded by video in Australia and Europe, including the four video records made from three observing sites for this partial eclipse.
Examination of each of the video records for variation in umbral geometry over the time of the eclipse, has shown that dynamic change in the size of the umbra can be detected by image analysis measures (Soulsby, 1994b).
The umbral size for this partial eclipse, corrected for refraction, topocentric declination and spherical projection as described in Appendix B, also exhibited change in semi-diameter as shown in Figure 9. For the Lavington and Ellenbrook image measurement plots, the mean values of F2 were 0.501 degrees and 0.474 degrees respectvely. Even though there is some difference in the values for the two sets of image measurements, a similar trend can be seen in each, illustrating that change did occur in the umbral size during the eclipse.
It is puzzling that all umbral measurements are low when compared with the values found by crater timing analysis. However, the umbral size for 1995 April 15 appears to be similar to that found for most other eclipses where video records were analysed by this technique.
Previously published F2 mean values (Soulsby, 1994b) for the lunar eclipses of 1989 February 20, 1990 February 9, 1990 August 6 and 1993 June 4, uncorrected for topocentric declination or spherical projection, all exhibited greater fluctuation throughout each eclipse and with mean values of F2 all generally lower than their theoretical umbral semi-diameters.
The crater timings provided gave discrete points on the umbral edge for a mean value of observed umbral enlargement and oblateness to be determined.
The crater timings produced a non-statistical observed umbral oblateness for this eclipse of 1/80 found from a linear best fit to the computed umbral semi-diameter (Ro) and position angle on the umbra (Ø ), as shown in Figure 10. The eight crater immersion timings made by Peter Skilton and 23 timings by Maurice Clark, also produced comparable theoretical umbral oblateness values of 1/123 and 1/176 respectively.
These values were supported by the primary contact times estimated from the video records which were well defined for both contacts in the Lavington and Ellenbrook images, and showed that the umbral oblateness during this eclipse was greater than the classical value.
The data provided by eight observers of this partial lunar eclipse have allowed estimates to be made of the geometry of the umbra. The values obtained from the video records show change in umbral semi-diameter with values much lower than the crater timing outputs. Peter Skilton's ten timings gave an observed umbral semi-diameter of 0.747 degrees while Maurice Clark's data gave a mean of 0.749 degrees for 23 crater immersion timings and 0.750 degrees for 39 emersions.
Measurements of captured frames from two of the four video records gave some variation in umbral size (F2), as expected due to the differing instrumentation. With corrections applied for refraction, topocentric declination and spherical projection (see Appendix B) these gave an overall trend as shown in Figure 9, demonstrating that changes in the transparency of the atmosphere of the Earth caused a gradual change in umbral size, which increased at mid-eclipse.
The data from the crater and primary contact timings indicated that the umbral oblateness increased, and the observed times reported by Robert Price for first and fourth primary contacts were effectively the same as both contact times derived from the author's ILEE theory. This was supported by the linear best fit to two sets of crater timing observations which gave an observed umbral oblateness of 1/80.
The consistent low values of umbral semi-diameter obtained from the video records are puzzling; however, the corrections used for this eclipse have reduced the fluctuations seen in previous analyses. Further work is needed to ascertain the cause of the low F2 values obtained from the video records - at least the records are available for further measurement, perhaps by other software.
The Calwell Lunar Observatory was well prepared for this eclipse with three optical systems (f/7 15 cm Newtonian telescope, 10X 50 binocular and a 400 mm telescopic lens with 2X adaptor) all on an electronically driven equatorial mount. Two video cameras (a monochrome CCD surveillance camera at 350X and a colour HandyCam at 70X) and one photographic camera (a Cannon T70 on the 800 mm effective focal length lens with polarising and UV filters). Video recording experience from previous eclipses and the author's continuing participation in the ALPO's Clementine Project (Soulsby and Brakel, 1994, 1995) was at hand, but alas observation was defeated by consistent heavy cloud on the evening of this eclipse.
While the weather for this eclipse was most disappointing, the worst cloud cover experienced by the author since commenced his program of lunar eclipse observation in 1972, the observations of Robert Price and his team at Bethanga, John Bennett at Lavington, Peter Skilton at Frankston, Brian Sture and Harry Moller at Ellenbrook, Graham Wolf in New Zealand and Maurice Clark in Perth, were all very welcome.
A vote of thanks is recorded to all for their input which prevented an unwanted "discontinuity" in my 23 year old programme. But happily, the 1995 April 15 lunar eclipse is now the 27 th in the Soulsby series of lunar eclipses.
It is planned to continue the southern hemisphere lunar eclipse programme, primarily with South America over the next two years (1996 April 4 - total, 1996 September 27 - total, 1997 March 24 - partial). However, it is with renewed interest that we look forward to the 1997 September 16 total eclipse, fully visible only in Western Australia and to the next two Australasian total lunar eclipses of 1999 July 28 and 2000 July 16.
Soulsby, B.W., 1990. Improved lunar eclipse ephemerides, J. Brit. astr. Ass., 100, (6): 293-305.
Soulsby, B.W., 1993. Analysis of the 1993 June 4 total lunar eclipse, Aust. J. Astr., 5, (3): 85-92.
Soulsby, B.W., 1994a. The Lunar Eclipse Observer, 1, (4), 11 pp.
Soulsby, B.W., 1994b. Change during lunar eclipses, in B.W. Soulsby (ed.), Proc. 16 th Natn. Aust. Conv. amat. Astrer., Canberra, pp 101-114.
Soulsby, B.W., 1994c. Four lunar eclipses from South America, Aust. J. Astr., 5, (5):167-176.
Soulsby, B.W., Brakel, A.T., 1994. The Clementine Project and transient lunar phenomena (TLPs), Aust. J. Astr., 5, (6): 201-209.
Soulsby, B.W., 1995. The Lunar Eclipse Observer, 1, (5), 11 pp.
Soulsby, B.W., Brakel, A.T., 1995. Clementine and transient lunar phenomena, Southern Sky, May/June 1995 : 50-54.