«Geological Mapping of Remote Mountainous Regions Using Metric Camera Imagery Initial Experiences with Photogrammetric Space Images *) By Manfred F. ...»
77 S. 115-149
Vienna, December 1984
Mitt, österr. geol. Ges. 1984 11 Fig., 5 Tab., 15 PI.
Geological Mapping of Remote Mountainous Regions Using
Metric Camera Imagery
Initial Experiences with Photogrammetric Space Images *)
By Manfred F. BUCHROITHNER
With 11 figures,
15 plates and 5 tables
The applicability of Metric Camera space imagery for geological reconnaissance
mapping and for more detailed studies up to a scale of 1 : 50 000 is demonstrated.
Three mountainous regions of different geological structure in western Saudi Arabia, in the Afghan Hindu Kush and in the East Nepalese Himalaya are stereoscopicly analysed using colour-infrared photographs. Detailed examples are given. The author summarizes the advantages and drawbacks of Metric Camera imagery for geological mapping in remote areas.
Resume L'applicabilite des images d'espaces faites par la Camera Metrique pour la cartographic geologique de reconnaissance et pour des recherches plus detailles jusque ä l'echelle 1 : 50000 est montree. A l'aide de photographies en coleurs infrarouges trois regions de montagne, differentes en structure geologique, en Arabie seoudite occidentale, ä I'Hindou Kouch en Afghanistan et ä I'Himalaya en Nepal d'Est sont evaluees stereoscopiquement. Des exemples detailles sont montres. Les avantages et les desavantages des photographies faites par la Camera Metrique pour des recherches geologiques dans des regions lointaines sont presentes.
Zusammenfassung Die Verwendbarkeit von Weltraumbildern der Metrischen Kamera für geologische Erkundungskartierungen und detailliertere Untersuchungen bis zum Maßstab 1 : 50 000 wird aufgezeigt. Anhand von Farb-Infrarot-Photographien werden drei im geologischen Aufbau unterschiedliche Berggebiete im westlichen Saudi-Arabien, im afghanischen Hindukusch und im ostnepalesischen Himalaya stereoskopisch Author's address: Univ.-Doz. Dr. Manfred F. BUCHROITHNER, Institute for Image Processing and Computer Graphics, Graz Research Center, Wastiangasse 6, A-8010 Graz, Austria.
*) Published within the scope of ESA Metric Camera Experiment "High Mountain Research in Southern Central Asia" and research project no. P 5668 of the Austrian Fonds zur Förderung der wissenschaftlichen Forschung (Austrian Science Foundation).
8* Manfred F. Buchroithner ausgewertet und Detailbeispiele gebracht. Die Vor- und Nachteile von Aufnahmen der Metrischen Kamer
1. Introduction Since the beginning of 1984 a new type of space image data is available: geometrically well-defined high-resolution space photography. Despite the existence of operational tools like Landsat MSS and TM data a present trend of emphasizing the importance of photogrammatric space photography can be obviously noticed. This development shows its clear expression in the Metric Camera and the Large Format Camera (LFC) Experiment with its associated Attitude Reference System (ARS) designed by the U. S. National Aeronautic and Space Administration (NASA). The possibility of better photogrammetric data collection and, hence, more effective topographic map production over large areas seems to be one of the major impulses for this new step in earth-oriented remote sensing.
1.1 General Remarks on the Metric Camera Experiment The first Spacelab mission of the European Space Agency (ESA) was launched on NASA's Space Shuttle flight No. 9 from Cape Canaveral in the USA on November 28, 1983, the Shuttle landing in Dryden, California, on December 7, 1983. This manned mission was jointly conducted by ESA providing the Spacelab and NASA offering the Shuttle flight. The mission carried 37 major experiments, one of them (no. 33) being the Photogrammetric (Metric) Camera Experiment (KONECNY & Geological Mapping of Remote Mountainous Regions 117 1979). It consisted of a Zeiss Aerial Survey Camera RMK 30/23 with a
SCHRÖDERfocal length of 305.28 mm mounted over the Spacelab window and operated from space at about 250 km nominal flight altitude with two film cassettes loaded with colour-infrared or black-and-white film. The High Quality Window (HQW) has been designed, constructed and tested by the Austrian Vereinigte Metallwerke Ranshofen-Berndorf Aktiengesellschaft (VMW), thus representing Austria's space hardware contribution to the Metric Camera Experiment on Spacelab 1. With the above mentioned film material on December 2, 3, 5 and 6, some 550 colour-infrared (table 1) and 470 black and white photographs were taken over China, Central Asia, the Middle East, Africa, Europe, North, Central and South America at an image scale of about 1 : 820,000 in single strips of 189 km width, at mostly with 60% (e. g.
Saudi Arabia, Hindu Kush), but partly also with 80% (e. g. Himalaya) longitudinal overlap.
Table 1: Spectral range and formation of the Metric Camera colour-infrared film positives (second generation set by Kodak Cibachrome method). In this diagram the colour reversal process is shown. Colour reversal transparencies are produced by developing the film as a dye image. Combinations of the dyes result in a positive image of the scene. (Drafted using written information by KODAK 1976.)
1.2 Mission Objectives of the Metric Camera Experiment The Photogrammetric Camera mission on Spacelab 1 brought several firsts, the most relevant one for the present study being the following: it is the first time that space photography is purposely suitable for stereoscopic data collection and standard photogrammetric instrumentation. Therefore it is possible to use the imagery
a) for mapping and map revision of planimetry (see section 5.2),
b) for measurement of heights, particularly for those taken with 80% overlap (see section 5.1), and
c) for thematic mapping in different disciplines (geology, hydrology, forestry, landuse) with the aid of stereo observation.
Manfred F. Buchroithner On Spacelab 1 within 4 operating hours, 11 million km2 of the earth's surface were photographed, 70% of which were suitable for mapping. This amounts to 5% of the land surface of the earth in a single experimental mission.
1.3 Metric Camera Imagery For the Spacelab 1 mission the Shuttle was launched with a 57° inclination, providing photography of areas up to 58° northern and southern latitude (KONEC
& SCHRÖDER 1979). However, the late time of launch, combined with the launch NY window of 16:00 h GMT at Cape Canaveral resulted in solar illumination at all photographed areas that was never more than 30° at the beginning of the mission and 10° at the end of the mission. This meant that the full scene exposures of /IOOO sec. had to be prolonged to /soo or Vim sec. resulting in image motions of up to 14 m or 28 m during the exposures. Furthermore, the low sun angle reduced image contrast (depending, of course, on relief and surface cover). Because of these unfavourable conditions, NASA agreed to refly the mission with the originally scheduled 3 film cassettes on the NASA Earth Observation Mission 1 (EOM-1), to be launched on June 12, 1985 (KONECNY 1984). Table 2 shows the Spacelab 1 mission performance and Photogrammetric Camera design data.
The first generation diapositives reserved for the ESA principal investigators were produced at the Deutsche Forschungs- und Versuchsanstalt für Luft- und Raumfahrt (DFVLR) at Oberpfaffenhofen. The black-and-white film was developed at DFVLR only 3 days after the landing of the Shuttle, and the colour-infrared film was developed at the Institute Geographique Nationale (IGN) at Creil 4 days after the landing (Table 2). The photographs treated in the present studies were submitted to the author together with over hundred offical Metric Camera experiments all over the world within the course of June 1984. This means a delay in data dissemination that is far beyond the originally planned "some weeks after the mission" (cf.
KONECNY, REYNOLDS & SCHRÖDER 1982); a fact that should change for further missions due to the requirements of fast information acquisition in certain fields.
Some of the early results for the geological applicability of Metric Camera data shall be given in this paper.
2. Methodical Remarks
2.1 General Notes The studies presented this paper were carried out within the ESA-accepted research project "High Mountain Research in Southern Central Asia (HMR in SCA). Experiment for the Metric Camera Experiment in the First Spacelab Payload (Earth-Oriented Mission). STS 9/Spacelab ESA. 28. November-8. December 1983" (except the Saudi Arabia scene). Principal Investigator of the above mentioned ESA Project is R. Kostka, Technical University of Graz.
Using imagery from three regions with completely different morphological, geological and cultural features, the applicability of photogrammetric camera images of mountainous areas for geological mapping was studied. This resulted in some initial reference data for forthcoming photogrammetric remote sensing images, such as the U.S.-initiated Large Format Camera (cf. KONECNY, REYNOLDS & SCHRÖDER 1982).
2.2 Procedures of Interpretation The geological interpretation of the 23 cm X 23 cm colour-infrared film positives from the Hindu Kush area and the Himalaya was carried out on a light table by means of a Wild ST4 mirror stereoscope using l x, 3x and 8X binoculars. These magnifications correspond to the scales 1 : 821 700, 1 : 273 900 and 1 : 102 700 120 Manfred F. Buchroithner respectively. Both scenes were evaluated in a 1 : 1 mode on a transparent overlay, except the various enlarged subscenes (fig. 2, 3, 8). Those were interpreted in the scales given in the respective figure captions. It is logical that for a relevant geological interpretation the three magnifications used had to be applied in an alternate, interactive way. As already mentioned before, except the Himalaya scene with 80% overlap, the other two scenes had 60% overlap in flight direction. Nevertheless small portions of the Saudi Arabia and Himalaya imagery were also analysed in areas where no stereoscopic effect was available. These monoscopic photo interpretations were admissible in areas with almost no relief, where one could easily extrapolate from the respective stereomodels (that is the coastal plains of the Red Sea in the Saudi Arabia scene and the Ganges plains in the Himalaya scene). In the first case, this value amounted to 40% of the area of a whole image frame (a big portion of it being occupied by sea) and to 20% of the southernmost Himalaya scene.
The areas shown in the tentative geological interpretation sketch maps (plates 2, 6,
10) of the various scenes are 21 350 km2 in Saudi Arabia, 35 570 km2 in the Hindu Kush area, and 62 630 km2 in the Chinese-Nepalese Himalayas.
The Saudi Arabia stereo-image pair was analysed directly at the graphic working station of the Institute for Image Processing and Computer Graphics of the Graz Research Center at a scale of approx. 1 : 820 000. Photographs and mirror stereoscope were mounted on a Summagrid digitizing light table. The linear interpretation elements and border lines of two-dimensional surface features were, instead of using a pencil, traced with a cursor. The resulting drawing could be visualized and checked on a Tektronix 4014 graphic screen, using the subsystems MDS, IDAD, and MAPOUT of the DESBOD software package, all developed at the above mentioned institute (RANZINGER, KAINZ & LEBERL 1981; KAINZ & RANZINGER 1983; BUCHROITHNER 1984; HÜTTER 1984; KAINZ, RANZINGER & HÜTTER 1984; BUCHROITNER in preparation). The result of this "digital interpretation" was then plotted on a Benson 1212 3-colour drum plotter, offering the opportunity to select almost any desired scale within a certain range.
The drawing of the profile line of the north-south cross section through the Himalayas (plate 11) was carried out on a Kern DSR-1 analytical stereo plotting system using the CRISP software package developed at the Institute for Image Processing and Computer Graphics of the Graz Research Center (FUCHS & LEBERL 1984). The same hard- and software facilities were also, besides the conventional method of using a Rost Type 644 planimeter, used for the interpretation and computation of the 1 : 50 000 erosion study in the upper Solu Khumbu area, Nepal (cf. fig. 7-9, plates 13 and 14, tab. 3; BUCHROITHNER 1984, BUCHROITHNER, in prep.).
In comparison with the interpretation of multispectral space data, Metric Camera imagery offers the advantage of using well-established and better known methods of "conventional" black-and-white and colour-infrared photo interpretation respectively.
Apart from the major terrain features, such as prominent peaks and major drainage lines/rivers, lake/sea shores, vegetation boundaries and settlements, the Geological Mapping of Remote Mountainous Regions major linear structural features (lineaments/faults/fractures) and two-dimensional surface features (photogeological units) were included into the geological sketch maps. For these interpretations, only major geological guide lines or in-depth data from comparatively small areas (say some 30 X 30 km2) were known. Intentionally, almost no detailed information on larger portions of the studied scenes was used to demonstrate to what degree of soundness and detail a first geological reconnaissance mapping can be achieved with only a small amount of collateral data in a comparatively short period of time. The esteemed reader and expert on the respective regions is invited to evaluate the validity and accuracy of the space photo mapping presented in this paper.
Knowledge from earlier trips passing through parts of the Hindu Kush and Himalaya region depicted on the Metric Camera scenes were quite beneficial for the interpretation. Further, a field study in an area few hundred km NW of the Medina scene, some 15 years ago, yieled a general understanding of this type of landscape.
3. Saudi Arabia Scene
3.1 Image Characteristics The Metric Camera image stereopair available from Saudi Arabia covers a portion of the Hejaz, the region between the city of Medina (Al Madinah) and the Red Sea (fig. 1, plate 1). It consists of the images 01-0120-03 and 01-0121-03, the first figure indicating the Spacelab mission, the second the number of the photograph within an operation strip (i. e. flight line) and the third one the operation strip number. The data were acquired on December 2, 1983, at 06-28-41 and 06-28-50 GTM, with a forward overlap of 60%.