NavList:

   

Reply

The reason for the difference has finally been located. However,
before explaining it, I first would like to answer George's questions:

> Another question relates to that 6x zoom. Is Marcel comparing the Sun and
> the landmarks on the same photo, so that we can be sure that the zoom level
> is the same for both? Or are they taken at different times, in which case we
> need to be convinced that setting the zoom to (presumably) its maximum of x6
> is reproducibly the same on each occasion.

No, those are different photos. In the case of the sun the scale is
derived statistically from measurements done in over hundred photos.
In the case of the landscape feature several photos have been made at
different days and from different locations (and even at different
times of the day). Nobody is perfect: yes it very rarely happens that
the zoom is not completely set up to the mechanical stop; this can
later easily been noticed since the size of the object is suddenly too
small. The different photos from the landscape feature compared well
with each other.

> Next is the angle subtended between two landmark points. Marcel tells us the
> spacing between them has been determined by Google. Does that correspond to
> old-fashioned paper mapping? How precisely has his own observation post been
> located, as that's a factor in the calculation?

Yes, the cause of the problem is here. Let me explain how it arose; it
will also show the type of difficulties one has to overcome at certain
places on this planet.

Since a long time I'm looking for a reference feature with known
height. For this I was looking for detailed maps, but couldn't find a
shop to buy them. It's only recently that I learned that it is illegal
in this country to sell such maps or even hand out copies of them. A
professor suggested that I indicate him the places of interest which
he then would look up for me. I'm still waiting for his answer.... In
this situation I started to look for a 'feature' where I could
estimate the height from relationships between horizontal and vertical
dimensions using GoogleEarth and GoogleMap for measuring the absolute
horizontal dimensions. I finally found a small island where several
private photos could be found on the Internet showing the island from
different directions; to this collection I added one of mine. A
governmental Internet page even mentioned its height as being 90m.
From the analysis of 6 photos resulted finally a height of 87m which
agreed quite well with the 'official' height. From my own photo, the
only one where the viewing angle was exactly known, resulted a height
of 86m. As a further verification I calculated now the height in my
photo also using my pixel scale derived from the sunset photos and the
distance measured with GoogleEarth and obtained only 79m. Looking also
at my other photos of this island showed that the scale is 182.6
pixels per degree compared to 198.2 as derived from the sunset photos.
The good agreement of the estimated height with the 'official' value
made me question what could be wrong with the scale obtained from the
sun.

Not finding a possible reason why the pixel scale derived from the
sunset photos would be so much wrong, I started finally to question
the measurements of the distances which were made with GoogleEarth. In
order to verify this I made in the meantime photos of a large building
in 1.7km distance which has distinct features and is also well visible
on GoogleEarth. Out of a careful analysis a scale resulted of 208.7
pixel per degree! This makes it now clear: measurements done with
GoogleEarth don't have the accuracy as the shown value may allude.
Btw: the mean value of the two scales obtained with GoogleEarth (182.6
and 208.7) is 195.7 this agrees already much better with the 198.2
(+/-1.5) as obtained from the sun.

Thank you, George, Greg and Bill, for your help. Your reflections and
suggestions have very much been appreciated.

Marcel

--~--~---------~--~----~------------~-------~--~----~
NavList message boards: www.fer3.com/arc
Or post by email to: NavList@fer3.com
To , email NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---

   

Reply