Classifying depth-layered geological structures on Landsat TM images by gravity data: a case study of the western slope of Songliao Basin, northeast China

Geological structures on remotely sensed images, Landsat Thematic Mapper (TM) images in this case, can be classified by quantitative depth information on the basis of the comparison of results from Landsat TM images and geophysical data. Although the lineaments with different depths can be visually interpreted together on Landsat TM images, the depth information and geological significance of these lineaments are however hard to obtain solely from the Landsat TM images of a study area under a thick cover, and it is of much importance for hydrocarbon exploration in the Western Slope Belt of Songliao Basin, northeast China. During the present study, the 3-dimensional field source information, including location and depth information, is derived from 3-dimensional Euler deconvolution of gravity data in particular. As an example, it may be quantitatively classified into four groups of depth range: < 100m, 100-500m, 500-1000m, > 1000m. It is then superimposed onto the lineaments map from Landsat TM images using a geographical information system (GIS). With a comprehensive analysis of the superimposed maps, we obtain validation and quantitative depth information of the geological structures delineated on the Landsat TM images. Four depth-layered maps of geological structures with different depths are presented here. It is concluded that the number of structures with depth greater than 1000m on the Landsat TM images is fewer than those at the other three depth ranges. The detection of geological structures on Landsat TM images attributed to depth information derived from the geophysical data may also be possible by this approach.