Perhaps the work done for NASA to draw hour and date lines on images
from the Mars rovers' sundials could be used to reverse engineer photos
of earthbound dials.
Woody Sullivan posted some background info on the MarsDial a few weeks
ago (see below). One of the URLs he recommended
(http://redrovergoestomars.org/marsdial/images.html) mentioned "image
analysis software":
"Student Astronaut Abby and Student Astronaut Shih-Han used
image
analysis software to determine that the center of the sphere's shadow
corresponds to a time on Mars at the Gusev Crater site of 11:01 AM on
sol number 11 of the mission."
Given the playful, but strict, nature of nature's laws, the reversing
engineer will not be able to extract enough info from a 2-dimensional
photo to completely solve the 3-dimensional problem posed by Geoffry
Lane. It helps greatly if some of the variables are known, for example
-- as John Carmichael notes -- the declination of the wall on which the
dial is mounted.
Perhaps the MarsDial image analysis software can be tortured into a
form that will at least take the drudgery out of cut-and-try
approximations by the reversing engineer?
Tom
Greetings:
The first extraterrestrial sundial now sits on a planet about
150 million km away. When it was fabricated here at the U. of
Washington in 1999, I held it in the sunlight falling on our planet,
and now it's being bathed by light from the same Sun, but at a
distance about 50% farther, and therefore an intensity about ~25% as
much. (Note also that the angular size of the Sun is ~20' as seen
from Mars, which makes the penumbrae of all shadows smaller, i.e.,
sharper-edged.)
We're still working out some of the bugs in the software to
accurately project a grid of hour lines and date lines on to the
image of the MarsDial as taken by the main camera on the mission
(Pancam). For instance, if you look at the posted images, the 12-hr
line does not appear to pass through the center of the base of the
gnomon, which it should for a level surface. The Rover was, however,
tilted by a few degrees while sitting on the lander (and of course
just yesterday it moved on to the Martian soil - hurrah!), in which
case the 12-hr line does *not* pass through the center of the base.
Our software does take care of any arbitrary tilt, but I'm still not
100% confident we've got the superposition right. Note also that
there are really two main cameras (for stereo imaging), and for some
reason one camera is giving different results than the other! Stay
tuned.
Cheers, Woody Sullivan
Here are some URL's of interest:
http://redrovergoestomars.org/marsdial/
- home page for MarsDial
project, with basic explanations (this text is now being improved and
expanded)
http://redrovergoestomars.org/marsdial/images.html
- actual images
from Mars of the MarsDial with hourlines and datelines
http://www.planetary.org/mars/tpr_marsdial.html
- detailed article
about the MarsDial and its development; originally published in "The
Planetary Report" (serious gnomonicists should start with this
article)
http://planetary.org/news/2004/mars-dial.html
- press release on
MarsDial (8 Jan. 2004)
http://pubs.giss.nasa.gov/abstracts/1997/Allison.html
= 1997
technical article (PDF) by M. Allison; for those who want **all** the
details and equations for time on Mars
******************************************************************
Prof. Woodruff T. Sullivan, III Center for Astrobiology & Early
Evolution
Dept. of Astronomy Box 351580
Univ. of Washington tel. 206-543-7773
Seattle, WA 98195 USA fax 206-685-0403
-
Dave Bell wrote:
One significant problem we have in reverse engineering many of these dials
is that we cannot be sure the original photo was taken straight-on, normal
to the dial face. Many of the photos submitted to John have been shot from
the ground or other accessible vantage point. In orthogonlizing the
images, errors are introduced, because the gnomon and dial face are not in
the same plane. I don't think these can easily be corrected.
-
|