2D STRATIGRAPHIC ANALYSIS FROM DOM USING SPATIAL TECHNIQUES

Stratigraphic and sedimentologic analysis are often based on data from 1D stratigraphic columns, conveying poor information about the lateral extension of rock units. We propose an approach based on Digital Outcrop Models (DOM) that combines field descriptions, 2D interpretation and data processing. DOMs provide the opportunity to model 3D geological bodies and structures but 2D analysis generates complementary information using standard, GIS-based spatial techniques on non-horizontal projections. Artificial exposures are better suited for bidimensional analysis because they provide less tridimensional information than their natural counterparts. Digital outcrop models (DOM) allow to generate high-quality orthomosaics with useful orientations that optimize the information that can be geologically interpreted.

We present two case studies that combine spatial and geologic techniques to efficiently produce stratigraphic and structural data on new exposures on a highway in northern Colombia (Lower Magdalena basin): turbidite depositional systems in shallow marine environments of the Paleocene San Cayetano Formation and shallow-marine to deltaic transition recorded in the Oligocene Cienaga de Oro Formation. Both outcrops offer the possibility of quantifying sedimentological, stratigraphic and structural processes on inaccessible exposures.

Area and thickness measurements, proximity, overlay and adjacency analysis, image segmentation, and linear referencing techniques where used along with lithofacial descriptions, photointerpretation and structural descriptions to test an approach to stratigraphic areal analysis. Linear referencing allowed to extract lithofacial information on lines crossing the 2D bodies. The possibility to automatically generate multiple stratigraphic columns provided a means to quantitative analyze changes in thickness and shape, and to produce single segment columns in contrast to field-based descriptions. Additionally, linear referencing and overlay functions were used in computing fracture abundance measurements that account for the effect of areas with patchy debris coverage on fracture detection.