International Journal of Applied Earth Observation and Geoinformation
Using airborne hyperspectral data to characterize the surface pH and mineralogy of pyrite mine tailings
Introduction
Acid mine drainage (AMD) is a key concern of mining industry due to its impact on the quality of water and soils surrounding mine waste deposits. Acid mine drainage derives from the oxidation of metal sulphides, e.g. pyrite (FeS2), exposed to oxygen and water (Singer and Stumm, 1970). Pyrite can be found within waste rock dumps, ore stock piles and tailings of many mines. During the production of AMD, iron and sulphur from pyrite are oxidized generating sulphuric acid leachate. The leachate acidity is capable of releasing heavy metals contained in the waste rock affecting water quality and leading to metal enrichment in sediments. This phenomenon is amplified by the rock crushing and grinding process because oxygen has greater access to the pyrite grains.
Reaction pathways from the oxidation of pyrite lead to the formation of a suite of secondary minerals (e.g. copiapite, jarosite, schwertmannite, goethite, and hematite) (Singer and Stumm, 1970, Kleinmann et al., 1981) stable under a range of pH conditions (Fig. 1). Secondary iron-bearing minerals present diagnostic absorption features at visible and short wave infrared wavelengths (0.4–2.5 μm) (Cloutis et al., 2006, Crowley et al., 2003, Montero et al., 2005) and are thus amenable to detection via imaging spectroscopy. The direct detection of pyrite from spectral observations is less evident than that of secondary iron-bearing minerals because pyrite displays low reflectance, saturated Fe-absorptions, and is typically coated by oxidation products (Swayze et al., 1996). Mineral mapping in mine tailings has been investigated through imaging spectroscopy for more than a decade (Swayze et al., 1996, Swayze et al., 2000, Farrand and Harsanyi, 1997, Ferrier, 1999, Lopez-Pamo et al., 1999, Dalton et al., 2000, Kemper and Sommer, 2002, Ong and Cudahy, 2002, Williams et al., 2002, Ong et al., 2003a, Ong et al., 2003b, Shang et al., 2002, Sares et al., 2004, Montero et al., 2005, Rockwell et al., 2005, Riaza et al., 2011a, Riaza et al., 2011b, Riaza et al., 2012). Riaza et al., 2011a, Riaza et al., 2011b, Riaza et al., 2012 undertook mineral detection from airborne and field spectral data for a suite of locations in the Iberian belt. They also undertook simulations of satellite data to assess the potential of mineral mapping with the upcoming EnMAP hyperspectral satellite system. A number of studies report a pattern in the spatial distribution of minerals in tailings namely a central zone of pyrite with low soil pH transitioning outwards to minerals formed in gradually more neutral pH conditions (i.e. hematite) (Montero et al., 2005, Rockwell et al., 1999, Rockwell et al., 2000, Rockwell et al., 2005, Swayze et al., 1996, Swayze et al., 2000). This zonation pattern suggests a decrease of potential acid mine drainage as a function of the distance from a pyrite rich ore zone.
Predicting tailings leachate pH using imaging spectroscopy is an emerging remote sensing application with limited studies having been realized by Ong and Cudahy (2002), Ong et al. (2003a), Sares et al. (2004), Williams et al. (2002) and Zabcic et al. (2009). Such a capability would supplement traditional methods (i.e. ground surveys) that are challenging to implement due to the extent and large volume of mine waste.
This study reports regional scale mineral maps for pyritic mine tailings generated from remote sensing spectral information to characterize tailings of the Sotiel-Migollas complex in Spain and pinpoints sources of AMD. We present a detailed investigation of four tailings from the Sotiel-Migollas complex, generating mineral maps from airborne hyperspectral imagery that are assessed using spectra from samples collected in the field and associated X-ray diffraction (XRD) measurements. In Zabcic et al. (2009) we used the same spectral imagery to generate maps of soil leachate pH. Thus here discuss these soil pH maps in the context of the mineral maps generated in this study along with the relative merits of both map products. Lastly, we examine spatial patterns in image maps that can be used to monitor the impact of mine waste over time, and help detect contamination in the surrounding environment through filtration, spillover or runoff that cannot always be detected on the ground. Such monitoring can support efforts of decision makers to understand the spatial extent of high-risk areas and prioritize possible remediation activities.
Section snippets
Study area
The Sotiel-Migollas mine complex is located in the vicinity of the towns of Sotiel-de-Coronada and Calanas in the region of Andévalo, province of Andalucia, South-West Spain (Fig. 2). The mine complex includes the processing plant, two tailings ponds, three major waste rock tailings, and various minor areas of rock waste or tailings.
The mine is situated in the Iberian Pyrite Belt (IPB), which is considered the largest volcanogenic massive sulphide (VMS) deposit in the world. The IPB extents for
Airborne spectral measurements
Airborne hyperspectral data (HyMap) were acquired on August 14, 2004 and June 17, 2005 to coincide with field campaigns. HyMap has 126 spectral bands covering the 0.45–2.48 μm infrared region with contiguous spectral coverage (except in the atmospheric water vapour bands), and full width at half maximum between 13 and 17 nm. The acquisition covered a maximum area of 97 km2 between Northing 4151550 and 4173316 and Easting 680906 and 696357 (Fig. 2). The pre-processing of the imagery included a
Extraction of image endmembers and labelling
The generation of final mineral maps was conducted with the June 2005 HyMap reflectance imagery. Pixels occupied by vegetation were first removed from the image by calculating the normalized difference vegetation index (NDVI) (Rouse et al., 1973) and using a minimum threshold of 0.3. Spectral endmembers were then extracted for the remaining image pixels using the spatial-spectral endmember extraction (SSEE) algorithm of Rogge et al. (2007). This algorithm was developed to improve the detection
Mineral interpretation derived from laboratory spectra of sediment samples
The mineralogy interpreted from laboratory spectra served to guide the interpretation of image endmembers extracted from airborne data. Table 2 reports the mineralogy interpreted from laboratory spectra of 78 samples. Most encompass mineral mixtures, since the mine tailings are intimate mixtures of minerals. Sixty seven samples have one or more dominant minerals (shown in bold) including hematite (18), goethite (11), pyrite (8), ferrihydrite (7), schwertmannite (7), and jarosite (5).
The
Predictive pH maps and links to mineral maps
In Zabcic et al. (2009) our laboratory investigation of pH predictive models based on samples led us to explore a pH predictive model using all 126 available HyMap bands. Applying this model to the validation set for the HyMap imagery gave a r2 of 0.71 and a RMSE of 0.77 (Zabcic et al., 2009) comparable to a model derived from the FieldSpec FR data.
Predictive pH maps obtained by applying this model to the HyMap imagery reveals very low pH values as seen in Fig. 5. Overall, the pH values
Conclusions
The extraction of spectral endmembers from imagery revealed 26 endmembers for tailings material that represent mostly mineral mixtures. From these, eleven spectral groups were defined, each encompassing minor variations in mineral mixtures. Each of the ten groups was dominated by one of the following minerals: copiapite, goethite, hematite, jarosite, halotrichite, pickeringite, pyrite, rozenite, schwertmannite, and szomolnokite. A distinctive mineral zonation pattern is observed around the
Acknowledgements
This research was supported by a discovery research grant to B. Rivard from the National Science and Engineering Research Council of Canada Discovery Grant 194260. HyMap imagery and field work was provided and supported by German Aerospace Center (DLR).
References (39)
- et al.
Detection and discrimination of sulfate minerals using reflectance spectroscopy
Icarus
(2006) - et al.
Mapping the distribution of mine tailings in the Coeur d’Alene River Valley, Idaho, through the use of a constrained energy minimization technique
Remote Sens. Environ.
(1997) Application of imaging spectrometer data in identifying environmental pollution caused by mining at Rodarquilar, Spain
Remote Sens. Environ.
(1999)- et al.
Secondary sulfate minerals associated with acid drainage in the eastern US: recycling of metals and acidity in surficial environments
Chem. Geol.
(2005) - et al.
The microbiology of acidic mine waters
Res. Microbiol.
(2003) - et al.
The extent of the Aznalcollar pyritic sludge spill and its effects on soils
Sci. Total Environ.
(1999) - et al.
Integration of spatial-spectral information for the improved extraction of endmembers
Remote Sens. Environ.
(2007) - et al.
Odiel River, acid mine drainage and current characterisation by means of univariate analysis
Environ. Int.
(2003) - et al.
Assessing mine drainage pH from the color and spectral reflectance of chemical precipitates
Appl. Geochem.
(2002) Automated MESMA Unmixing for Fractional Cover Estimates
(2007)