Article: Discrimination of the geological and structural units of the precambrian basement of Western Africa by means of multispectral transformations: case of the square degree of Korhogo in the North of Côte d’Ivoire

Koffi KOUAME 1, Jean Patrice JOURDA 2, Mahaman Bachir SALEY 2, Miessan Germain ADJA 3 et Fernand Koffi KOUAME 2

1 Centre Universitaire de Recherche et d’Application en Télédétection (CURAT), UFR des Sciences de la Terre et des Ressources Minières de l’Université Félix Houphouët Boigny de Cocody
2 Centre Universitaire de Recherche et d’Application en Télédétection (CURAT), UFR des Sciences de la Terre et des Ressources Minières de l’Université Félix Houphouët Boigny de Cocody
3 Département des Sciences et Technologies, Section des Sciences de la Vie et de la Terre

Abstract

Mineral prospecting represents a broad scope of geological research, and development of mineral resources. In Côte d’Ivoire, the mineral exploration suffers from a weak production of geological maps (Anonymous, 2012). The remote sensing, thanks to the diversity of the produced images and the progress in the processing and the GIS, constitutes for the geologist of exploration a tool of investigation that must allow to fill this insufficiency. However, because of the plant cover and the lateritic the profits obtained with the classic processing are often mitigated. To overcome this obstacle to a better geologic mapping, new techniques of image processing, such as the multisensors merger, the radar interferometry were developed. The present study explores the multispectral transformations in Korhogo region.
The study area is Korhogo Square Degree. It lies between latitudes 9 ° and 10 ° North and longitudes 5 ° and 6 ° West (Plate 1). Based on a granite or schist bedrock, soils are mainly lateritic ferruginous tropical interwined (Sedes, 1965). This region is characterized by a NNE -SSW succession of bands of schistose rocks, plutonic and migmatitic (Plate 2) (Arnould, 1963; Delor and al., 1995).
In the present study, the multispectral transformations, specially the Selective Principal Components Analysis (SPCA) and the bands ratios associated with the directional weighted Sobel 7x7 filters were applied to the Landsat 7 ETM + images of Korhogo region.
At lithological level, multispectral transformations consisted mainly in Selective principal components analyzes (SPCA) and the calculation of band ratios. Ratios method based on the notion of reflectance is the ratio between the amount of radiation reflected by a material and the amount of energy received by the radiometer for a given wavelength (Scanvic, 1996). Choosing the index is guided by the nature and the spectral behavior of the object to identify or simply by the objective in satellite images (Youan Ta et al, 2008.). Thus, reports of classic bands (ETM +4 / ETM +5; ETM +4 / ETM + 6, ETM +5 / ETM 6, 7 ETM / ETM 6) were made to mitigate the effects of shading related to relief and the incident light. Different SPCA conducted with Landsat 7 ETM + bands and principal components obtained are listed in table 2. For naming components, subscript numbers indicate the bands in the SPCA.
At the structural level, the application of directional filters Sobel helped highlight image discontinuities equivalent to lineaments. They are sometimes associated with alignments of biological indices (termite alignment or plant species) or geomorphological discontinuities (mineralogical variations in armor or soil) (Koussoubé, 1996). Plate 3 summarizes all the treatments performed.
Extraction of the information contained in images resulting from different treatments to improve contrast and enhancement the semantic contents of the images was done only by visual interpretation. In this approach, the internal information of the image (spectral radiance and relative positioning or neighborhood) is described through the analysis of various parameters, including tone, color and texture. The image in Plate 6a allows us to distinguish different colors (gray, white and dark) and two shapes (circular, rectangular) plots of Ferké sugar complex. Texture refers to the visual effect that produces a surface with a local variation of gray tones, but nevertheless suggesting a homogeneous appearance on the whole. Consideration of several adjacent pixels to set the property texture introduced in fact absent in the case of luminance alone spatial dimension. Caloz and Collet (2001) state that “the remote sensing image is seen in this way, as a raw material, combined with the other information and know-how, will feed the model of reality”.
Major lithological units in the region have been delineated through visual interpretation of several colorful compositions and other products of specific treatments. These include the first principal component of the SPCA of the bands 5 and 7 (Plate 8), colorful compositions of the principal components of the SPCA of the bands 1, 2 and 3 (Plate 9) on one hand, and of the bands 4, 5 and 7, (Plate 10) on the other hand, and the ratio ETM+5 / ETM +4 (Plate 11).
The combination of all data interpretation has used to draw a sketch of the remote sensing lithological map of Korhogo Square Degree (Plate 12). This map identifies five (05) major geological set oriented SW-NE:
– The Niellé massive that occupies the northwest corner of the map;
– The Dioula Series which extends to Tongon resort in the Southwest;
– The Korhogo massive that occupies the entire middle portion of the map, leaving always appear in its central part, the circumscribed granite Korhogo;
– The High Bandama Series between the towns of Komboro and Ferké;
– And the Ferké massive to the East, covering the regions of Tafiré (South -East) and Ferké.
Visual interpretation of neo-channels generated from the application of directional filters 7x7 Sobel helped develop a lineament map (Plate 19). Structurally, some features and shear corridors were mapped with a dominant SW -NE orientation.
The lithological and structural information’s of images obtained from spectral transformations and gradients Sobel has used to produce a lithostructural recognition map of the Precambrian basement in Korhogo Square Degree (Plate 24). Figure 25 and 26 show the correlation between these results and those of the geological map and the map of mining indices from previous work (Delor et al, 1995. Sodemi 1996; Jourda et al, 2006.) With sharper boundaries. It appears many similarities with previous work accidents. Geological boundaries are sometimes sharper. In addition, a metamorphic aureole nonexistent in previous works has been mapped.

KEYWORDS
Selective Principal Component Analysis, bands ratios, lithostructural mapping, Korhogo, Precambrian basement, Côte d’Ivoire.