653 - 
nt measure- 
asurement 
ves more in 
tensity lines 
j values thus 
iy the photo- 
3 density dif— 
n the absolute 
68). The 
>to in such a 
;h one of which 
l completely 
e oldest, so- 
mlsion and its 
£ht for some 
3 (1967) used 
\nother proce- 
md positive 
>r with a slight 
lit to repro- 
>ional film, 
ble results, 
n using a 
irmines the 
ity range of 
[ 3rd order can 
sodensitracer 
Quantimet 720 
;is for the 
Lt. It clearly 
hological pur- 
Photo 1 
Vertical photograph of a drowned Karst area in Florida, USA, 
used for densitometric analysis and showing the dry, 
intermediate, and wet zones mentioned in the text. 
The image used for the purpose is a vertical aerial photograph at a scale of 1 : 
20, 000 of a drowned karst area with numerous sinkholes in Florida, USA (photo 
1). A fairly dry area occurs in the upperleft corner where the sinkholes are 
largely covered by grass and other low vegetation. Their density contrast with 
the non-sinkhole area is only slight. Stereoscopic relief analysis adds to their 
identification. A wet zone occupies the lower right portion of the image. The 
sinkholes are mostly filled with water and are then darktoned. In the intermediate 
zone that stretches diagonally across the image the sinkholes run dry after the 
wet season and the barren and possibly somewhat saline lake bottoms are light- 
toned. The image thus clearly illustrates how very different the tonal rendering 
of one and the same type of terrain can be. It is obvious that direct density 
measurements are pointless. An interpretation-overlay had therefore to be made, 
on which all sinkholes were marked in black. The overlay was thereafter used for 
quantitative study. Time consuming interpretation and drawing was thus required 
and the quantitative results ultimately obtained are affected by interpretation bias 
and are thus not completely objective. The limitations of density analysis for the 
study of terrain is thus clearly brought to light.