Full text: New perspectives to save cultural heritage

CIPA 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
the aerial photos (1:15000) and by the resolution of the scanner 
(15pm). The colours of the infra-red images were converted by 
means of a colour transformation- and intensity matrix, with 
additional parameters of practical experience into images of 
natural colouration*. 
Again this map, including the revised photogrammetric line 
analysis and the city map, the latter generalised to a scale 
1:5000, resulted in a GIS (Geographic Information System). 
Contour line, inclination and shaded relief maps were derived 
from the terrain model. 
One aim was to create a layer of the archaeological resources 
for the orthophoto map scaled 1:5000. For that purpose the 
digital city map of Ephesos had to be generalised. This map was 
generated by surveying and by digitising existing paper plans 
(scales between 1:50 and 1:200). The plans were tied up 
through measurements of control points in the field; then the 
digitised plans were converted by a neighbourhood 
transformation (Hardy, 1972) (multiquadratic interpolation); 
thereby an optimal absolute infitting under maintenance of 
inner geometry of the drawings was obtained. Figure 2 shows 
the area of the East gymnasium with generalised vectordata and 
the underlying orthophoto in natural colouration. 
Infra-red pictures are often used for photo interpretation work, 
because of their special properties. The emulsion of the film is 
sensitive to the spectrum of near infra-red light and represents it 
in red colour. The vegetation emits strong infra-red rays and 
therefore allows to distinguish between different plants. 
Because of the strong influence of the ground conditions on the 
vitality of the plants, it is possible to see them at the colour 
gradation in the photos. 
To verify the aerial photo interpretation an area south of the 
Arkadiane was chosen, which is a colonnaded street leading 
from the theatre to the habour. Here the Byzantine city wall (in 
Figure 3, drawn in blue) encloses a rectangular area* **. 
Archaeological excavations took place: at the gate on the 
southside of the Arkadiane, at the street leading south, at the 
westgate of the Agora, at the so called gate "Medusentor" and at 
the street leading east to west. On the aerial photograph a 
rectangular place is visible, showing a circular monument in its 
centre, the place is surrounded by a pillared hall. The square 
measures on its sides 85 m and the diameter of the central 
circular building is to 20 m. 
5. WATER CONDUIT 
In ancient times Ephesos was supplied by several great water 
conduits. One of them starts in the south at a place called 
Degirmendere, it has a total length of 43 km (Ozi?, Atalay, 
1999). 
Starting point for the following computations was the 
consideration that ancient water conduits often follow the 
topography. There is the question, if it is possible to model 
approximatly the course of the water conduits by applying a 
mathematically defined terrain surface of DTM. If it is 
Carried out by Prof. Dr. Dipl.-Ing. J. Jansa, Institute of 
Photogrammetry and Remote Sensing, TU-Vienna. 
** F. Hueber, Ephesos (1997). Gebaute Geschichte (1997) pp. 
51.: "Der Platz südlich des Viersäulenbaues ist heute an der 
byzantinischen Stadtmauer zu erkennen, die seine 
Außenwände als Teil der Befestigung mitbenutzte. Sonst 
sind von diesem Platz nur noch einige Säulenbasen der 
Kolonnaden erhalten." 
succesful, it could optimize fieldwork and model areas where 
no architectural remains are left. 
In this case an elevation of 62 m on the aerial analysis was 
messured at the aqueduct Arab-dere-kemer, which is the nearest 
to Ephesos. By following the calculated run of the water 
conduit in the direction to Ephesos on the DTM and by 
assuming a fall of 0,5 Promille*** for the calculations, the 
course as shown in Figure 4 comes out. This results corresponds 
well with an elevation of 60 meter over the sea proved by 
Forchheimer (1923) the water conduits on the north side of the 
Bülbül Dag. 
5.1 Errors 
Which errors can occure by simplification? 
a) By the calculation every valley is driven in the full 
length without any aqueduct, therefore a greater distance 
has to be covered and more difference of altitude "is used". 
b) b) Also a terrain saddle, which in ancient times was 
used to shorten the distance by digging a canal or a tunnel, 
therefore construction expense was reduced. 
Both errors are model errors, which are leading to the 
assumption of a water conduit lying too deep in the area of 
interest. 
What are consequences of errors in altitude necessarily lead on 
the model, and what are the consequences of changes of the 
landscape on the determination of the position of the water 
conduit? 
c) c) In a very steep area of 50% inclination as it occurs 
on the north slope of the Bülbül Dag, an altitude error of 
±lm and an error of position of ±2 m has to be calculated. 
d) d) If you have now an inclination of 10%, 1 m would 
amount to an uncertainty of the position of ±10m. 
e) e) Having an even area, it is not possible to estimate 
the line course only in presumtion of an incline. 
Those cases show very clear by the importance of the 
stochastically model for DTM, to be able to value the reliability 
of the results. For the model calculations in this place discussed 
only the starting point and the terrainmodell were assumed. If 
there could be introduced more points, that define the water 
conduit in the field and are surveyed, the calculations could be 
improved. 
REFERENCENS 
Forchheimer, Ph., 1923. Wasserleitungen, In: Forschungen in 
Ephesos Band III, Wien, pp 233. 
Hardy, R.L., 1972. Geodetic applications of multiquadratic 
analysis. Allgemeine Vermessungsnachrichten 79, Herbert 
Wichmann Verlag, Karlsruhe, pp. 398-406. 
Hueber, F. 1997. Gebaute Geschichte, von Zabern, Mainz am 
Rhein. 
Özi§ Ü., Atalay A., 1999: Fernwasserleitungen von Ephesos, In: 
H. Friesinger - F. Krinzinger (Hrsg.), 100 Jahre Österreichische 
Forschungen in Ephesos. Akten des Symposions, Wien 1995, pp 
407. 
*** Ozi§, Atalay (1999) prove a incline of 5 per mill, this shoud 
be an errat, because my estimation of the elevation 
difference from the spring to the aqueduct including the 
length of the water conduit results a value of about 0,5 per 
mill.
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.