977
nt developed
ng in 10 or
in a number
al needs are
bands only,
ines have a
on in orbit-
>f cost: one
ing only the
we need and
ds therefore
ose applying
ies and en-
d the space
jcause it is
required for
considerably
governments
applications
reach in the
:e and other
stored in a
and retrie-
ed to accom-
le surveying
ormation for
iguently, we
ssible. This
is our goal
bed informa-
a world-wide
ral types of
.ng simulta-
ds of appli-
several de-
;he way, il-
on resource
few decades
:e a hole in
ale aircraft
ises surveys
agraphs made
,f they were
Le, acquisi-
systerns com-
1 multispec-
tels for re-
s Earth sur-
md/or other
is ion of the
allites were
present con-
non-stereo-
on of 80 or
natch for a
¡solution of
.1 certainly
of resource
or thematic
000.
;es are the
ites. Their
parts of the
ition render
Llion) scale
oceanography
ly fields of
\A, GOES and
al and will
lunch of the
of the sys-
icluding the
r temporari-
ti surface in
There may be scope for a third type of resource
satellite, catering for resource mapping at scales
in the order of 1:250,000 to 1:500,000. While main
taining the multispectral and stereoscopic capaci
ties of the first type already mentioned, lower
spatial resolution and the coverage of a larger
surface area per frame should be aimed at, possibly
in combination with a higher temporal resolution.
Thorough exploration of the market for such products
should precede future developments in this area.
In addition to these three main types, satellites
serving specific areas may prove viable, such as the
Tropical Earth Resources (TERS) satellite concept
studied by Indonesia and The Netherlands for appli
cations in equatorial countries.
In the overview on current trends in remote sen
sing and programmes underway, the Earth Observation
Programme of the European Space Agency should be
mentioned. During the opening session European co
operation in the field of remote sensing from space
has been highlighted as a paramount example of suc
cessful international programmes. European countries
together with Canada decided to develop the ERS-1
programme. The objectives are to transfer experimen
tal use of microwave techniques to operational use
for applications in global ocean and ice monitoring.
This programme offers a challenge for both global
climate research, sea state forecasting and near
future operational and commercial applications. In
this respect data continuity is an essential condi
tion. After ERS-1 (to be launched in 1990) with an
expected lifetime of three years, the duration of
the mission will be expended by the launch of ERS-2
in 1993.
The ESA Earth Observation Preparatory Programme
which started this year will define future missions
dedicated to meteorology, land observation, ocean
and ice observation and solid earth studies.
With regard to land observation a major contribu
tion is expected from the acquisition of synthetic
aperture data on a regular basis. The ERS-1 program
me will offer a first opportunity to collect SAR
images in C-band overland, with the constraints
related to regional coverage over Europe and North
America and its experimental character. A possible
future land-oriented mission can provide usin the
90-ties with a capability for frequent observation
by means of an advanced SAR system suited to develop
operational use of polar orbiting platforms for
monitoring of agriculture and land use.
PHOTOGRAMMETRY IN THE SPACE AGE
Although the imagery provided by the first gener
ation satellites has to some extent put also to use
in small-scale topographic mapping and for map re
vision, most photogrammetrists have shown only a
limited interest in the imagery produced because the
limited spatial resolution and the absence of
stereoscopic relief rendered this material unsuit
able for most of their work. Mention should be made,
however, of the mapping of (parts of) the moon
making use of the stereoscopy provided first by the
libration of that celestial body and later by lunar
mapping systems such as on lunar orbiter and Apollo.
The situation has changed drastically in recent
years with the availability of high-resolution
stereoscopic pictures of the metric camera, and
large format camera experiments. Stereoscopy for
them is even more crucial than for earth scientists.
The question is: will the photogrammetrists evolve
into "imagegrammetrists" as photo-interpreters in
past years became image interpreters? Doubtless the
field of photogrammetry is changing and widening
with the introduction of new kinds of imagery. The
use of radar imagery restitution and of digital
cameras in close range applications are good exam
ple. Corrections for Earth curvature became impor
tant when satellite imagery came into use. Although
the photogrammetric potential of SPOT imagery and
of future satellites of this type is undeniable, it
is likely that also in the future a substantial
part of large scale (e.g. cadastral) photogram
metric mapping will be done on the basis of photo
graphic imagery using aircraft as a platform. The
answer to the question thus is —at least in part—
"no", even though digital analysis techniques will
increasingly be applied.
An interesting change in the relationships be
tween photogrammetric and resource surveyors is
noteworthy, however: The common interest of both in
the past was aerial photography with the resource
surveyor gradually devoting part of his attention
to satellite imagery. The field of common interest
in the future will be high-resolution satellite
imagery. In addition, the image interpreter will
use low-resolution imagery and may occasionally
look at an aerial photograph for very detailed
work. The photogrammetrist will continue to pay a
substantial part of his attention to aerial
photography, as will be explained below.
The increase in spatial resolution of satellite
imagery has been spectacularly rapid. For continu
ously operating systems the following data can be
given:
Meteo satellites 1960 > 1000 m
Landsat MSS 1972 80 m
Landsat TM 1982 30 m
SPOT-1 1986 10/20 m
Non-continuously operating optical systems based on
recoverable film or electro-optical systems based on
recoverable tapes (MOMS) perform even better:
Soyuz/Salyut Zeiss Jena MKF-6 camera
(20 m photographic) 1970 8m pixel
Metric camera (20 m photographic)
1983 8 m pixel
Large format camera (NASA/Itek)
1984 5 m pixel
MOMS (line array) FRG 1983 20 m pixel
MEOSS FRG 1987? 50 m pixel
Metric camera reflight 1990 3 m pixel
equivalent
equivalent
equivalent
equivalent
equivalent
equivalent
The spatial detail visible on these photographic
images is high. The fact that these systems func
tion only at irregular intervals is a drawback,
not so much for photogrammetry and cartography, but
particularly for resource surveying and monitoring.
The vertical accuracy of satellite images has
also risen to levels that become photogrammetrical-
ly interesting. Where occasional Landsat images
provide stereo in the sidelap zone of adjacent
passes, the stereo threshold is at best in the
order of 100 meters. Experiments with the metric
camera and the LFC have shown that, where good
ground control is available, approximately 15-20
meters vertical precision is possible. In the ab
sence of adequate ground control, 40 meters is the
maximum obtainable accuracy. For SPOT, a stereo
capacity of 5 meters is claimed in height and 3 m
in planimetry.
Three main facts emerge from these data:
1. Satellite imagery suits the needs for photo
grammetric mapping in scales of 1:100,000 and in
the near future even 1:50,000.
2. The high resolution obtainable by long focal
length cameras, particularly when operating from
lower altitudes (space shuttle: 250-300 km)
compared to orbiting satellites (SPOT:
approximately 800 km), make optical (photographic)
systems an interesting alternative not withstanding
the lack of continuity.
3. The second generation of satellite images do
not meet the high horizontal/vertical precision
required for detailed photogrammetic mapping car
ried out for, e.g., cadastral work. Aerial photo
graphy remains a necessity in this area.
Some photogrammetrists at present still have