GRSG Conference 2022: Orbit to Outcrop
Title: Identifying Copper Hot Spots in the Subsurface of an Operational Tailing’s Facility Using Landsat Time-series Analysis.
Author: Yaron Ogen
Space-borne remote sensing has established itself as an excellent and reliable tool for environmental monitoring due to its advantage in rapidly and consistently mapping broad areas. However, the obvious disadvantage of remote sensing in the VNIR-SWIR regions is that it can only collect spectral information on the surface and thus cannot be utilized to measure, monitor, or evaluate the subsurface zone.
However, with free and open access to the vast databases spanning five decades of multispectral spaceborne imagery and the recent development of high-speed processing interfaces, remote sensing technology can now be exploited for research of the subsurface domain using time-series analysis. This is especially valuable in areas where anthropogenic activity has resulted in an accumulation of material over a long period of time, such as in tailing areas. Chemical and textural analysis of today’s surface material indicates that the highest copper concentrations are associated with specific grain-size fractions (mainly sand) that are predominantly found in the dry areas of the tailing.
This information will be used to determine the spatial distribution of copper hot spots in the historical Landsat imagery. Thus, the main objective is to reconstruct the historical development of an active tailing basin associated with a porphyry copper deposit in Erdenet, Mongolia, using remote sensing data to assess and map its copper content. For that, we employed three datasets to conduct the research: drill core samples, Landsat images, and historical data of the composition of the tailings after the flotation process.
In December 2020, eleven drill cores were extracted by the Erdenet Mining Company from selected locations in the tailing area, with depths varying from 20 to 88 meters below the surface. Samples were collected every 2-3 meters, which summed up to a total of 281 subsurface samples. Additionally, the copper content of each sample was measured using an inductively coupled plasma mass spectrometer (ICP-MS) in the Central Geological Laboratory (CGL) in Ulan Baatar, Mongolia.
The time-series analysis was performed using Google Earth Engine with combined surface reflectance collections of Landsat 5 (1985–2012) and Landsat 8 (2013–2020) between May and September (snow free period), which covers most of the operational period of the tailing facility. Each image in the collection has been subjected to initial processing, including factor correction and cloud masking. Following that, we used a variety of spectral indices, like the normalized difference water index (NDWI), to assess the historical surface conditions that might be associated with the presence of high copper content.
Our preliminary analysis of the data indicates that the copper content spilled into the tailings is decreasing primarily due to improvements in the ore processing method, while minor changes in content are due to the concentration of copper in the ore itself. Results from the drilling samples indicate that, in general, the concentration of copper increases with depth. However, there are strong vertical as well as horizontal variations in copper content due to the discharge technology and micromorphology of the tailing basin.
This study presents preliminary results regarding the exploration of copper residues in the subsurface of a tailing area using time-series analysis of Landsat images. Focusing on exploiting secondary raw materials may reduce material and energy consumption, reduce the human impact on the environment, and may even reduce production costs. However, it is necessary to improve the prediction accuracy.
To do so, our future work will combine grain size analysis with examination of proxy variables and include additional spectral indices that may indicate the presence of high copper concentrations in the subsurface domain. Subsequently, if the developed method yields satisfactory results, it can be applied to other tailing areas or even integrated with other space-borne sensors to identify copper and other economically valuable minerals.