Mineralogical mapping of the upper Benue trough Northeastern Nigeria from multispectral remotely sensed data using Linear Spectral Unmixing

The geologic formation of the Gongola arm of the upper Benue trough in Northeastern Nigeria is highly heterogeneous and conglomerate that discrete classification of its satellite image will not give the required detail information about its mineralogical constituent. In this work, a fuzzy classification algorithm, the Linear Spectral Unmixing (LSU) was used to classify the image with the aim of producing detailed mineralogical map of the formation. The mineralogical units of the formation were grouped into three: sanstone/Shale/limestone, Shist/Gneiss/Migmatide and Basalt. Fractional maps, showing the relative abundance of these major units were produced. The results showed the effectiveness of LSU, a fuzzy classification technique, as the best alternative in detail mapping of a highly conglomerate and heterogeneous formation.