Exploration for Emeralds Using Satellite Remote Sensing Imagery
The Coscuez mine in the Colombian Andes is an important source of emeralds, with a recorded history stretching back more than 400 years and production even earlier, to before the Spanish conquest. Located 100 kms north of Bogotá, the mountainous terrain is lushly vegetated as is clear in Google Earth, below.
This poses problems for visible/near infrared [VNIR] and shortwave infrared [SWIR] remote sensing systems, which sense only the top millimeter of the surface. Longwave infrared [LWIR] is better positioned, as thanks to thermal emissivity, some penetration of vegetation is possible.
A further issue for all optical remote sensing is cloud cover, as clouds and cloud shadows obscure the response of rocks and minerals.
L-band satellite synthetic aperture radar [SAR] can see through both clouds and vegetation and once the multipolarized reflections have been inverted to dielectric constants [DCs], we see Coscuez as a DC anomaly in a SAR image collected in January 2007. This anomaly is most likely due to the presence of pyrite in the deposit.
Several targets immediately jump out.
An Aster thermal image collected on 1 September 2017 is cloudfree over the mine. Five channels of LWIR data were unmixed into the sum of up to 16 unknown mineral endmembers using a proprietary spectral interpretation algorithm. These spectral endmembers were interpreted by comparing the image derived spectra to a spectral library measured in a laboratory. Three minerals dominate over the open pit and are interpreted as albite, kaolinite and beryl.
The spectral abundance for albite is a useful exploration targeting vector.
Coincident albite and conductivity anomalies should be considered as targets to follow up with field checking. Albite on its own produces too many targets, but these can be hi-graded using the coincident distribution of beryl as well as DCs estimated from SAR imagery.