CIPA 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
500 
CARL ZEISS, ck = 60 mm, format 9 cm ■ 12 cm) and 35 
images of a bundle block taken in 2002 (camera PENTAX67, 
ck = 105 mm, format 6 cm • 7 cm). 
: ' 
Figure 13: Surface of the dome inside, visualization of laser 
measurements 
A lot of natural points in the painted ornamental decoration of 
the dome could be defined and measured in both data sets, 
normally each point was identified and measured in more than 5 
images of each dataset. The measurements of the image 
coordinates were performed with the analytical 
photogrammetric stereo system WILD AC3. In the year 1979 
some points in the ornament were connected with a three 
dimensional geodetic network with an overall precision of 
±3mm. 
Figure 14. Deformation vectors of the dome, xy-plane, 
grid: 1 m, Cracks (A) and (B) signed 
The 1979-bundle block with a precision of ±2 mm was in the 
free adjustment settled with s-transformation to the geodetic 
points. The precision of the 2002-bundle block was the same: 
±2mm. 
After extensive common analysis of the two blocks and looking 
at the real situation of the cracks some points west of the cracks 
in the lower and physically less disturbed regions of the dome 
were chosen for the connection of the two blocks with the help 
of the s-transformation in the free adjustment. Three 
dimensional deformation vectors were projected to the xy- 
plane, see figure 14, and to the xz-plane, see figure 15. It is not 
surprising that the part of the dome in the east between the two 
discussed cracks (A) and (B) had completely moved about 1 cm 
in direction of the apse before the whole apse seems to be out of 
balance. The height of the dome had sagged about 2 cm maybe 
as a consequence of the horizontal moving of a part of the dome 
but the definite analysis will just be possible with some more 
information to be find in the future. From static point of view 
the cracks - especially (A) and (B) - are very important 
elements for a definite analysis. The question of the time 
linearity of the moving is unanswered , next time we will look 
for a third bundle-block of the year 1998, one year before the 
big Izmit - earthquake. 
— 
1 i i 
. i j.. 
V 
)} 
_L 
u 
’if 
4 
. 
\ 
1 1 
/j 
l| 
J 
1 
li 
( 
S' 
\ 
Jl 
\ w 
y 
i 
f-rn 
J 
s 
' ; T J* 
i 
/ 
\ 
3, 
- 
• : 
T 
N 
; ^ 
* 
■ -1 
7* 
! 
CO 
l i 
V *. 
u 
r 
"A’ 
r' 
s• 
a 
rrt 
t 
' 
f * 
v\ 
1 
. 
l i 
\ 
. 
t 
. 
i 
“7 
• 
. 
L 
r 
„ 
• i 
• 
; C 
m 
. 
I 
U ; 
Figure 15. Deformation vectors of the dome, xz-plane, 
grid: 1 m 
5. DOCUMENTATION OF THE EARTHQUAKE 
DAMAGES 
The shocks of the last big so called Izmit-earthquake in Turkey 
(in august 1999, magnitude 7,4) gave reason for some 
additional damages in the KAM which can be documented with 
photographs and photogrammetric images, look at the examples 
in figure 16. Some of the damages are really new but most of 
them are hidden before or placed at the weak points of the 
building well known to the experts, in figure 17 we see an 
example of that theory: with the help of digital image 
processing it could be shown, that the defect of the piece broken 
in 1999 was known some years before, the piece was once more 
renovated in former times.