289 
; data were able 
on. 
* composition of 
crease has been 
:ing in a rise in 
hat this global 
; effort is shown 
luction areas of 
iture would rise 
; parameters of 
is a complex 
research issues, 
global satellite 
!S are concerned 
the traditional 
)hotogrammetry 
e of information 
nvironment and 
en regional and 
; GIS serves as 
atic data from a 
from thematic 
world mapping 
) 1:200 000 we 
‘ 1:200 000 in 
on through the 
t the 1:50 000 
lopment analog 
the land areas, 
1/3 of the land 
ition effort in 
ill required for 
inents of Latin 
i. 
e update rate of 
e the 1:25 000 
1:50 000 map, 
r ears old. This 
iditional aerial 
mapping is not able to meet global updating 
demands. 
On the other hand satellite mapping offers a 
clear alternative to meet future demands. 
But the situation is also complex. The early 
satellite remote sensing efforts concentrated 
on the low resolution/high repetitivity potent 
ial of meteorological satellites and on the 
medium geometric and temporal resolution of 
resource satellites such as Landsat and Spot. 
With military restrictions being eased there is 
now the hope of using the potential of high 
resolution satellites for mapping. 
While the traditional governmental systems of 
the USA (Landsat), France (Spot), Japan 
(JERS-1), ESA (ERS), Canada (Radarsat) still 
concentrate on medium resolution, Russia, 
India and Germany has at least opened 
possibilities in the 2 to 5 m ground pixel 
range in preparation for the future. 
Looking at the resolution requirements it is 
clearly the 1 m ground pixel which is required 
for urban studies and for disaster monitoring. 
Governmental efforts in the planning stages 
with Landsat 7 and Envisat will not change 
this situation much, except in India, where 
Cartosat is aimed at 1 m pixels, in France, 
where Spot 5 aims at the 3 m pixel, in Japan, 
where ALOS aims at the 2.5 m pixel. 
But long anticipated advances are expected by 
the commercial systems by EarthWatch 
(Quick Bird), Orbital Sciences (OrbView 3), 
Space Imaging (Ikonos), and West Indian 
Space Ltd. (EROS B), all aiming at the 1 m 
pixel during 1999. 
To judge the potential of these upcoming 
images the existing images of Russian space 
photography (the KVR 1000, the KFA 3000) 
at 2 m pixel, the German MOMS scanner at 5 
m pixel and the Indian IRS 1C and D at 6 m 
pixel are available in samples. 
Regarding the cartographic potential this 
imagery is well suitable to provide the 
required planimetric accuracy, provided that 
control via GPS is locally available. 
Elevation accuracy still presents a problem if 
elevation contours of denser than 20 m are to be 
provided. 
With respect to object detectability a 2 m ground 
pixel is required for standard mapping. If such a 
pixel size is not available quality deterioration 
will result. 
Urban photomapping at the scale 1:10 000 is 
possible from 2 m Russian camera images of the 
KVR 1000. 
To solve the elevation requirements one should 
remember that optical stereo-measurements 
depend on the base-height ratio, which for most 
satellite systems is poor. 
To meet a 1 m contour requirement a pointing 
accuracy of ± 2 dm is required. This is 
unachievable by satellite systems. 
The Stereo-MOMS principle of on-line stereo 
can at best correspond to a ± 5 m point accuracy, 
suitable for 20 m contours. Stereo-MOMS with 
5 m pixels is, however, suitable for 1:25 000 
image maps, particularly when from these 
updates of maps should be made. Stereo-MOMS 
offers the added advantage to generate 
perspective views of the terrain. 
MOMS 02-P from MIR is still an operational 
sensor. Sofar over 100 M km 2 have been covered 
during its operation time. Even though 
processing is slow, since the tapes have to be 
brought down to Kazakhstan, and transported to 
Germany for processing a wealth of data has 
become available. 
Another 6 m pixel possibility is IRS 1C & D 
which through a DLR-GAF cooperation with 
India can be received daily in Neustrelitz, 
Germany, for Europe. 
Even though the European ESA-ERS 1/2 
program to be continued with ENVISAT does 
not provide the highest resolution, imaging radar 
is a cheap source of supplementary data. 
■ It can be used for multitemporal land use of 
studies, 
■ and it can be used for radar interferometry. 
Much has been promised for this technique, but 
relatively few independent investigations of the