Classifying airborne radiometry data with Agglomerative Hierarchical Clustering: A tool for geological mapping in context of rainforest (French Guiana)
Résumé
In highly weathered environments, it is crucial that geological maps provide information concerning both the regolith and the bedrock, for societal needs, such as land-use, mineral or water resources management. Often, geologists are facing the challenge of upgrading existing maps, as relevant information concerning weathering processes and pedogenesis is currently missing. In rugged areas in particular, where access to the field is difficult, ground observations are sparsely available, and need therefore to be complemented using methods based on remotely sensed data.
For this purpose, we discuss the use of Agglomerative Hierarchical Clustering (AHC) on eU, K and eTh airborne gamma-ray spectrometry grids. The AHC process allows primarily to segment the geophysical maps into zones having coherent U, K and Th contents. The analysis of these contents are discussed in terms of geochemical signature for lithological attribution of classes, as well as the use of a dendrogram, which gives indications on the hierarchical relations between classes.
Unsupervised classification maps resulting from AHC can be considered as spatial models of the distribution of the radioelement content in surface and sub-surface formations. The source of gamma rays emanating from the ground is primarily related to the geochemistry of the bedrock and secondarily to modifications of the radioelement distribution by weathering and other secondary mechanisms, such as mobilisation by wind or water. The interpretation of the obtained predictive classified maps, their U, K, Th contents, and the dendrogram, in light of available geological knowledge, allows to separate signatures related to regolith and solid geology. Consequently, classification maps can be integrated within a GIS environment and used by the geologist as a support for mapping bedrock lithologies and their alteration.
We illustrate the AHC classification method in the region of Cayenne using high-resolution airborne radiometric data acquired in 1996 across most of French Guiana. Access to the field in this region, almost entirely covered by tropical rainforest, is difficult, and therefore, use of airborne geophysical data is highly suitable to complement ground observations. Despite the vegetal cover, the U, K and Th maps, exhibit to the first order, a fairly good correlation with lithological units recognised by geologists in the field.
Consequently, classification of the radiometric data is globally concordant with existing 1:100,000-scale geological map. In addition, using the radioelement contents and relations between classes at different levels of classification, additional relevant information concerning weathering effects, unexpected lithological differences or transfer by erosion, are evidenced.