GISdevelopment.net ---> Application ---> Geology

Subcontinental Geologic Mapping using the MMRS Sensor (SAC-C Satellite Mission-Matutine Constellation). - An example of central regions of South America

Roberto Torra
Roberto Torra
Departamento de Geociencias.
Facultad de Ingeniería
Universidad Nacional del Nordeste. Güemes st. No. 749.
H3500CBO, Resistencia, Chaco, Argentina.
E-mail: roberto_torra@arnet.com.ar



Introduction
This paper is focused in describe and interpret the main geologic features of about 500.000 km2 located at the central South America region. The study is mainly based on the use of remote sensing low resolution -throught with a large area coverage- satellital data and techniques.

The study area comprises a fair rectangle outlined between 22º 00’ - 33º 00’ South Latitude and 54º 00’ - 62º 00’ West Longitude.

The study area embrace all the northeastern sector of Argentina, the west of Uruguay, the south, west and east of Paraguay and some neighbouring small sectors belonging to Bolivia and Brazil (see Figure 1).


The study was carried out working on SAC - C Satellite Mission images raw data named as path 223/000, 224/000, 225/000, 226/000, 227/000, 228/000 and 229/000 (orbits) which belong to the multispectral medium resolution sensor (MMRS). This is a 125 meters on-board optical-pixel resolution scanner-instrument.

Techniques of digital image processing were used to handle the BSQ satellital acquired data. Therefore we were able to create the corresponding scene compositions, mosaicking and cutting them and so delimited all the study area by means of computational procedures.

The final composicion used was 4/5/3 which adapted adecuately to conspicuous physiographical conditions on the wet seasson (summer and automn). Anywere we too used a 3/4/5 composicion in order to compare with 4/5/3 and so be able to extrapolate some very well known trainning areas of wide typical cobertures. These cobertures were perfectly checked out into the field and learn us many keys to later interpretation.

The clouded areas were, in general, minus than 5-10 in percentage on the entire mosaick (small sectors located at the extreme north and south of the builded mosaick).

Few studies were maked in regard with the South American continent physical and geological main elements present into the specific literature and concerning with extensive large areas of this area. However, many semi-regional studies exists in all countries here involucred. This reason let us to intend to summarize a logical comprehension design of some geologic phenomena within a regional context. That is to say, we begun a preliminary synthesis by means of the use of remote sensing as a tool for helful investigations. However, many years of field work seve us for the support of several of presents interpretations.

Because its very low costs (SAC - C digital optical data), to present days result imperative use satelitary images which could enourmosly help us into regional until subcontinental studies and analyses conducting to quickly and precise interpretations. This is a primary steep to overcome the study and analyses of the reasonable land use and land management. Then, we be able to create logical proxies for long sustainable development.

Methods
The SAC - C raw digital imagery CD disks data were supplied by the CONAE (Comisión Nacional de Actividades Espaciales de la Argentina, after an A/O agreement with the Department of Applied Geosciences, Engineery Faculty, National University of Northeastern (UNNE), Argentina) and then we used them for this work. The SAC - C paths (orbits) employed here are: 223/000, 224/000, 225/000, 226/000, 227/000, 228/000 and 229/000. The lighthing hour of ilumination was 10:45 a.m. Argentina time which correspond to 14:45 p.m. West Greenwich Meridian. The date of acquisition of all digital data were summer and automn 2002, Austral Hemisphere.

We analyze the matrix structure of the digital data images. Then, we cut aproximately a half image per path. This is because a lot of cloud presence overlies the Amazonas region (Brazil) and the extreme south of Argentina. The extension of the scenes employed herein, which cover all South America, also were taken in consideration because the size of the files were extremely large in size to handle into ATX PCs standards (more than 1,2 Gb per file).

After this, we create false color images for all the scenes and then mosaicking. Composition 4/5/3 and 3/4/5 were maked for geological, tectonical, morphological and vegetate coberture analyses. It is convenient to be present that the pixel dimension of these images is about 125 meters and so, it is not represent great precision in assembage. But on the contrary, they give us a very aproximate idea of the contextual regional main patterns. On the other hand, we have present the focus of our work which was cualitative and not quantitative.

The automatized classification were done used standard procedures as standard IDRISI® modules. We performed unsupervided and supervised tests. Because the amplitude involucred in this work, we categorize seven wide classes of cobertures as follow: vigorous vegetation (forest, woods, etc.), scrub and mixed grassland terrains (including intertropical sabannas with aloctonous palms), water masses (wetlands, lagoons, lakes, large rivers and it fluvial wide plains, etc.), several soils without organic coberture and few level of moisture (sometimes deforestation terrains and incipient desertification), soils with active agriculture but without discrimination among species and period of grouth, outcrops of recent Quaternary basat-alkaline volcanic aparatous, antropic constructions (urban, peri-urban and large cities).

Better results were obtained with supervided clasification although when the operators are well familiarized with previous results and so be able to configure adequately algoritms into the computational modules, the automatic classifications offers very similar results those of supervised classifications.

First orders geological, tectonical and morphological elements were outlined by means of in-screen digitalization using standard applications (software) as AutoCAD®.

Results
Main large crustal lineaments, not previously mentioned and never properly understood, were carefully detected and draw. The keys used for mapping lineaments were shadows and highlights and abrupt change of cobertures which implies a drastic variation into the tone-area and so lithology. The analyses of the main physiographical features as river drainage, low ranges, extense plains, large lagoons models, immense wetland masses models, deltas models, curvilines scarpments models, several extense faulting lineaments models were employed for interpretation. The present of some volcanic aparatous dome-model of Quaternary age were used too as an important tool for interpretation. The recognition of volcanic dome have two very different aspects: at the selvatic region they are covered by vigorous vegetation and so they give strong reddish colors apart from its typical geoform. At the areas with scarce vegetation they are circular and extremely fractured (see Figure 2). The wetland areas are characterized by dark tones whereas the agriculture areas present typical geometrical patterns as well as urban, peri-urban and large cities. The soils with low moisture and scarce vegetation are always brilliant. Incipient erosion processes are very well recognized in the head water drainage gullies, destroying silt materials (see Figure 3). The vigorous woods are brown dark tones and strong in texture, which sometimes is lobulated and so characteristic (Morello, 1968, Morelo and Adamoli, 1974).



The geological and tectonical first orders lineaments are clearly recognizable because its continuous mark into the composition scenes. Many times they separe different tonal color and texture designs, specially towards the east of the study area (e.g. Paraguay and Brazil).

Discussion
Upon the obtained results with SAC - C satelitary images they should be considered as a primary tool for large area investigations, specially of regional magnitude. It is clear that the 125 m pixel is not well pointed out for quantitative studies but it has a good chance within cualitative earth resources and ecologic researchs. Geological and tectonic lineaments as well as large units of vegetate coberture, soils, net drainages, incipient desertification processes, land and management use, morphophysiographical studies, all may be easy outlined with these digital optic low costs data (see Figure 4).

In comparison with Landsat series images they have some adventage because their possible regional vision in only one scene but Landsat series has a tremendous better pixel size dimension (30 m). However, we employed Landsat series images in some areas with more concrete interests (e.g. urban sites). The regional contextual physiograpy is better overcome by means of the use of SAC - C images. It is well known that when we need improve resolution with Landast images generally we quickly go to canal P SPOT (or IRS-C Indian Satellite digital data) images to create merge images and be able to improvement the geometrical and spatial resolution.

As a rule, it is seem very importat to define the objectives of our previous goals in order to select the appropriate images by the work (Chuvieco, 2000, Sabins, 1998 and others).

In this brief study, we can detected the main physical topics present at the region which let us better possibilities to understand and organize earth and ecology projects conducting to more exhaustive and detailed studies.

Conclusions
The SAC - C satellite images are a valuable tool for regional earth or ecology studies. This is so because we quickly can obtain a general view of the study in question. Then, we can handle information real and completely actual of the principal physical characteristics of the region. This was the case that occured in this preliminary study belong to the central South America region.

Another point for remark is the low cost of the study by means of the SAC - C satelitary images. It is very important to be present that with very few images per region, sometimes only one scene, we can get too much information about unknown territories.

The outlined physiographical characteristics of this area correspond to an extensional faulting system arrangement, which could be properly relationed to a subcontinental active partially simetrical immense rift system. The assemblage of delimited fault systems may be brokening the crustal surficial rock segments all along more than 2,000 km in a north-south direction. The pattern shows too a fracturing model of giant semi-circular rupture (colapse and sliping faults) to the east, at the Brazilian and Argentinian plateau and plains (Figure 4).

To the west area seems to be a little bit more straight faults. The oblique system faults also may be recognized, specially within the Paraguay and Argentine regions. The recent Quaternary vulcanism activity phenomena are surprisingly well developed at the Paraguay pampas (west). Near Argentina borders they appears as a “dome” model in the south-west Paraguay, to pass through Paleozoic and Mesozoic units. Many Pliocene-Quaternary volcanic aparatous appears at the proximity and within them of the Argentinian “Pampas Ranges”. Some domes are very well fractured which is easy to improve with directional filters.

It is proposed that this rift system could be extended and conected with the Amazonas river basin and finish at their mouth. In this case of very extended areas, the contextual consideration of the rift system analized turn on to very complex geological systems run into very different stages of evolution.

Apart from the imagery study, this conclusion is supported by more than ten years of field work as well as many sismic and wells register check data (Torra, 2001).

The final reflection concern that with a few economical (software and hardware) data and equipment we can be able to determine many surficial important earth and ecological features and so help people to generate sustainable ways . Better use of the land, quick monitoring of ecological and natural risk impact over human, floodings, masses landslides, etc. are among many aspects in which optical satellite SAC - C imagery can too much help us.

References
  • CHUVIECO, E. (2002) Fundamentos de Teledetección Espacial. 3rd Edición revisada. Editorial RIALP. Madrid. 567 p.
  • DELLEUR, J. W. (1998) Handbook of Groundwater Engineering. Springer-Verlag. Berlín. 967 p.
  • EASTMAN, R. J. (1997) IDRISI® for Windows. User’s Guide. Clark Labs for Cartographic Technology and Geographic Analysis. Idrisi Production. Clark University. 18 capítulos.
  • EASTMAN, R. J. (1997) IDRISI® for Windows. Tutorial Exercises. Version 2.0. Clark Labs for Cartographic Technology and Geographic Analysis. Idrisi Production. Clark University. 18 capítulos.
  • FOODY, G. M. (1999) The Continioum Proccesses of Classification Fuzziness in Thematic Mapping. Photogrammetric Engineering & Remote Sensing. 65 (4): 443-451.
  • GIRAUT, M., SOLDANO, A. and LUDUEÑA, S. (2000) Cuantificación de la evolución de coberturas hídricas durante las recientes inundaciones en el nordeste de la República de Argentina: Una propuesta metodológica. Dirección Nacional de Política Hídrica. Subsecretaría de Recursos Hídricos. Secretaría de Recursos Naturales y Desarrollo Sustentable. Buenos Aires. Unpublished report. 15 p. Argentina
  • JENSEN, J. (2000) Remote Sensing of the Environment: An Earth Resource Perspective. Prentice Hall Editions. New York. 544 p.
  • GÓMEZ, J. C. (1998) Interpretación Visual de Imágenes Aplicada a Catastro Rural. Dirección Provincial de Catastro. Geomática de Argentina S.A. Provincia del Chaco. Unpublished report. 85 p. Argentina
  • LILLELAND, T. M. and KIEFER, R. W. (2000) Remote Sensing and Image Interpretation. 4th Edition. John Wiley & Sons, Inc. New York. 724 p.
  • MILLER, R. (2000) Ecología y Medio Ambiente. Grupo Editorial Iberoamérica. Madrid. 652 p.
  • MORELLO, J. (1968) La Vegetación de la República Argentina. Las Grandes Unidades de Vegetación y Ambiente del Chaco Argentino. Primera Parte:
  • Objetivos y Metodología. Centro Nacional de Investigaciones Agropecuarias. Instituto de Botánica Agrícola. Serie Fitogeográfica Nº 8. Buenos Aires. 124 p.
  • MORELLO, J. and ADAMOLI, J. (1974) La Vegetación de la República Argentina. Las Grandes Unidades de Vegetación y Ambiente del Chaco Argentino. Segunda Parte: Vegetación y Ambiente de la Provincia del Chaco. Centro Nacional de Investigaciones Agropecuarias. Instituto de Botánica Agrícola. Serie Fitogeográfica Nº 13. Buenos Aires. 124 p.
  • PINILLA RUIZ, C (1995) Elementos de Teledetección. Editorial RA-MA Ediciones. Madrid. 315 p.
  • RYERSON, R. A. (1999) The Manual of Remote Sensing: Remote sensing for the Earth Sciences. 3rd Edition. John Wiley & Sons, Ind. New York. 701 p.
  • SABINS, F. F. (1998) Remote Sensing: Principles and Interpretations. W. H. Freeman and Company. New York. 493 p.
  • SISTEMA DE INFORMACION TERRITORIAL. (1991). Colección Territorio. Generalitat Valenciana. Conselleria de Obres Publiques, Urbanisme i Transports. Valencia. 49 p.
  • TORRA, R. (2001) Sedimentología de la Formación Ituzaingó (Mioceno medio), en los sectores norte y oeste de la Mesopotamia, Argentina. Unpublished Doctoral Thesis. Facultad de Ciencias Naturales e Instituto Miguel Lillo. Instituto de Estratigrafía y Geología Sedimentaria Global (IESGLO). Universidad Nacional de Tucumán. II Volumes. 417 p.
  • TORRA, R., SISTI, J.M., MURISENGO, W.G., GÓMEZ, J.C. and COLANTONI, L. (2000) Procesamiento de Imágenes y Cartografía Digital. XI Congreso Nacional de Fotogrametría y Ciencias Afines. CD ROM Actas. Santiago del Estero. Argentina
© GISdevelopment.net. All rights reserved.