Abstract Two recent papers, “Utility of high-altitude infrared spectral data in mineral exploration: Application to northern Patagonia Mountains, Arizona,” by Berger et al. (2003), and “Mapping hydrothermally altered rocks at Cuprite, Nevada, using the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), a new satellite-imaging system,” by Rowan et al. (2003), make a distinctive mark on the use of airborne and satellite hyperspectral imaging as an exploration tool. These two papers deal with imaging of the Earth’s surface using the visible (0.4 μ m) to near infrared (2.5 μ m) part of the electromagnetic spectrum to map various mineral species. Depending on their structure and molecular bonding, minerals reflect and absorb the electromagnetic spectrum in unique ways. A large group of minerals have distinct electromagnetic signatures that make it possible to identify them from imaging systems that map the range of the electromagnetic spectrum between 0.5 and 2.5 μ m. These papers represent two distinct approaches. The first paper, by Berger et al., discusses the use of the AVIRIS (Airborne Visible Infrared Imaging Spectrometer) scanner, which provides high-resolution reflectance measurements in the spectral domain (224 channels between 0.4 and 2.45 μ m) and variable spatial resolution (20 m), dependent on aircraft altitude. The second paper, by Rowan et al., discusses the use of the ASTER satellite scanner, which offers a limited range of spectra at three spatial resolutions (15, 30, and 90 m). ASTER measures reflectance radiation in 3 bands within the 0.52- to 0.86- μ m range (visible-near-infrared) at 15-m spatial resolution, and 6 bands between

Abstract The proliferation of remote sensing platforms has resulted in unprecedented opportunities for ore deposit vectoring. Importantly, remote sensing technology is now beyond the vague identification of alteration, and can accurately map specific minerals and directly contribute to the understanding of ore systems. Remote sensing is making discoveries of new alteration zones within classic and previously well mapped ore systems, as well as outlining their geometry and mineralogy. Confining this review to the geologically important reflected-light remote sensing systems, there are four main categories of sensors readily available to economic geologists, including the following: (1) submeter resolution panchromatic satellites that offer little spectral information but provide base maps; (2) multispectral Landsat satellites that can map iron and clay alteration; (3) the new ASTER satellite that can map important alteration groups and some specific minerals; and (4) hyperspectral airborne scanners that can provide maps of specific mineral species important to detailed alteration mapping. At the core of comprehending this plethora of technology is the difference between spectral and spatial resolution. This review will provide an understanding of the more fundamental aspects of remote sensing systems that will help field geologists to interact better with and leverage this rapidly evolving technology.

Abstract Alteration mineral assemblages are important to the understanding of and exploration for hydrothermal ore deposits. Conventional mapping tools may not identify fine-grained minerals or define important compositional variations. Field portable shortwave infrared (SWIR) spectrometers solve some of these problems and provide a valuable tool for evaluating the distribution of alteration assemblages. Spectrometers such as the PIMA-II allow rapid identification of minerals and mineral-specific variations at a field base. Mineral assemblages, integrated with other exploration data, are then used to target drill holes and guide regional exploration programs. Data collection must be systematically organized and carried out by a trained operator. Analysis of data sets requires the use of spectral reference libraries from different geological environments and may be aided in some cases by computer data processing packages. Integration of results with field observations, petrography, and X-ray diffraction analysis is necessary for complete evaluation. The PIMA (portable infrared mineral analyzer) has been used successfully in the high-sulfidation epithermal, low-sulfidation epithermal, volcanogenic massive sulfide (VMS) and intrusion-related environments. Case studies from these systems demonstrate the ability to rapidly acquire and process SWIR data and produce drill logs and maps. The resulting information is critical for targeting.

Abstract The determination of alteration mineralogy with field portable short wavelength infrared (SWIR) spectrometers is common in numerous exploration and mining programs, but frequently the application is simply one of mineral identification. In fact, such spectrometers and the recently developed core scanners can be used to map both mineral assemblages and changes in mineral composition, as well as provide data to aid lithological characterization and the development of alteration maps and three-dimensional deposit models. This paper utilizes data sets from three volcanogenic massive sulfide (VMS) deposits (Kristineberg, Sweden; Hellyer, Tasmania; and Teutonic Bore, Western Australia) to assess relative amounts of chlorite and white mica, determine subtle variations in the phyllosilicate compositions, and integrate voluminous data collected with a drill core scanning system (the HyLogger™). The data sets illustrate the variety of approaches to collection and analysis of field spectral data that are reasonable in the context of mineral exploration. These techniques may be applied in a wide variety of ore-deposit environments.

Abstract The physics of thermal infrared aerospace measurements is based on Planck’s Radiation law, Wien’s Displacement law, and Kirchoff’s law. The electromagnetic spectrum for thermal infrared aerospace measurements includes measurements beyond the reflected short- (2.5 μ m) to the long-wave infrared (14 μ m). Thermal infrared sensors measure thermal emission from the Earth’s surface in single wavelength bands (broadband), tens of bands (multiband), and in hundreds of bands (hyperspectral). Broadband thermal infrared measurement techniques include surface temperature mapping and thermal inertia mapping. Multiband and hyperspectral techniques involve mapping of changes in thermal emission at different wavelengths (emissivity mapping). Today, broadband surface temperature mapping is mostly done with satellite sensors. Thermal inertia mapping is done using broadband measurements taken during the day and night. Emissivity mapping is done using tens to hundreds of bands, and it requires sensors capable of measuring small changes in radiant emittance. Sensor systems discussed in this study include: Thermal Infrared Multispectral Scanner (TIMS), the Moderate Resolution Imaging Spectroradiometer (MODIS), Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Simulator (MASTER), the Spatially Enhanced Broadband Array Spectrograph System (SEBASS) and the ASTER satellite sensor. Several areas of Nevada, such as Brady’s Hot Springs, Steamboat Springs, Geiger Grade, and Virginia City, were used as sites for demonstrating the geologic applications of thermal infrared remote sensing. Corrected day and night images over Steamboat Springs were acquired by TIMS. These day-night images were combined together to produce a final processed temperature image, in which the temperature effects of albedo, topographic slope, and thermal inertia were minimized to facilitate the detection of geothermal anomalies. Spectral variations in emitted thermal energy were detected over the Geiger Grade and Virginia City areas using the MODIS-ASTER Simulator (MASTER) and (SEBASS). MASTER thermal infrared image data allowed two primary mineralogic units in the Steamboat Springs area to be identified: sinter and/or chalcedony deposits and quartz-alunite alteration, which have spectral emissivity features around 9.0 μ m; and clay-rich soil and clay alteration, which have spectral emissivity features around 9.7 μ m. The higher spatial and spectral resolution SEBASS data allowed six different alteration assemblages to be identified: quartz, alunite, pyrophyllite, feldspar, kaolinite, and montmorillonite and/or illite.

Abstract Landsat Thematic Mapper (TM)/Enhanced Thematic Mapper Plus (ETM+) has been known as the workhorse of geologic remote sensing because of its flexible spectral configuration, spanning through the visible, near-infrared, short-wave infrared, and thermal infrared portions of the electromagnetic spectrum, coupled with adequate spatial resolutions, world-wide coverage, and easy data access. Landsat TM-type sensors have produced a considerable amount of data since 1982, allowing an impressive achievement of reliable geologic and mineral exploration results. The purpose of this paper is to review, from the mineral exploration perspective, some of the historical developments that led to Landsat TM/ETM+, the fundamentals of the sensor, its imagery, and the processing techniques that have been employed for mineral exploration applications over this nearly 25-year period, since the first TM images became available. This paper also reviews some case studies in order to demonstrate the benefits of using the existing Landsat TM/ETM+ image database for mineral exploration, particularly in the new exploration frontiers. Although the future of the Landsat program and its data-continuity policy is not clear, mineral mapping applications based on existing data will continue because of its time and/or date independency. This signifies that historical Landsat archives will continue to be employed by future generations of remote sensing exploration geologists.

Abstract The emergence of hyperspectral image data to remotely map minerals on the surface of the Earth is the result of decades of laboratory, engineering, and physics research conducted by both commercial and government sectors. We present a synopsis of major events which culminated in the development of mature hyperspectral reflectance and thermal emissivity data collection capabilities that are currently available in worldwide capacity. Laboratory and field measurement efforts in the 1960s produced a foundation of knowledge pertaining to mineral spectral phenomenology. This in turn spurred the creation of first broadband multispectral and then hyperspectral data via airborne and spaceborne sensors to capitalize on mineral characteristics. Highly accurate radiometric calibration and atmospheric corrections allow remote measurements to closely emulate ground or laboratory measurements, while commercial hyperspectral data processing software enables timely analysis of the data. Increasing amounts of image data have been collected by a variety of commercial and government sensors, of which mineral exploration is a predominant application. The differences in sensor and image data acquisition parameters directly affect the ability to map minerals. A brief analysis of Cuprite, Nevada, using multiple sensors is used as an example to illuminate factors of importance.

Abstract Over the last decade, a new exploration frontier has opened up in Patagonia, Argentina, for gold mineralization. Important deposits, such as Cerro Vanguardia and Esquel, were discovered during this period. Despite these discoveries, basic geologic information for the vast Patagonia territory is still markedly scarce for generating targets for mineral exploration. In such a scenario, in which a prevailing arid climate and sparse vegetation often leave favorable bedrock exposed, multispectral satellite remote sensing is playing an important role in gold prospecting. This paper presents the results of a remote sensing project developed over northern Patagonia with the purpose of mapping potential areas of epithermal gold mineralization. A reconnaissance program using Landsat Thematic Mapper (TM) imagery to search for favorable targets on a regional scale was carried out, focusing on the mapping of possible hydrothermal alteration zones. The effectiveness of the results is evidenced by the fact that the majority of the hydrothermally altered targets pinpointed using Landsat TM showed anomalously high gold grades after follow-up geochemical analysis. Within the scenes analyzed, the region of Los Menucos stood out as having a high number of conspicuous altered areas carrying gold values greater than 0.1 g/t (some greater than 1 g/t). With the launch of Terra in 1999, carrying onboard the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), the Los Menucos region was further investigated to map alteration. A suite of advanced image processing methods was applied to the data, with the purpose of detecting and identifying key alteration minerals. The results accomplished in the Los Menucos region with Landsat TM and ASTER data demonstrate the capabilities of multispectral remote sensing for alteration mapping in frontier exploration regions, ranging from regional scale to district and deposit scales.

Abstract Recent advances in geometallurgy have made nickel laterites a premier target for mining and exploration companies. Parallel to this, a series of advances in remote sensing have become available at very low cost. Two of these advances, namely better topographical data, through NASA’s Shuttle Radar Topographic Mission (SRTM), and multispectral imagery from the ASTER sensor, have been combined to aid nickel laterite exploration in central Brazil. The Conceição do Araguaia region, located in Pará and extending into Tocantins state, is part of the Neoproterozoic Araguaia fold belt. The target area covers the Quatipuru mafic-ultramafic association, which includes serpentinites (metaperidotites and metadunites), talc schists, tremolite-actinolite schists and small volumes of pillow basalts, phyllites, BIFs, gabbroic and jasperoid rocks. These are enclosed regionally by slate to phyllitic rocks. Several other occurrences of mafic-ultramafic rocks, along with Quatipuru mafic-ultramafic association are interpreted as part of an ophiolitic complex. Low-grade (greenschist) regional metamorphism is dominant. Laterization has been active since the early Tertiary, resulting in an extensive regolith cover over the older rock units. In the present work, the Quatipuru mafic-ultramafic association was examined for nickel laterite mineralization by data compilation and remotely sensed image processing and interpretation. ASTER’s multispectral visible-shortwave infrared (SWIR) remote sensing capabilities were used to map areas of prospective mineral alteration and key mineral groups. Using spectral libraries for selected nickel-bearing minerals as standards, SWIR and visible-NIR bands of georeferenced mosaiced ASTER scenes were processed by feature-oriented principal component analysis (PCA), and the results converted in mineral-abundance images, based on statistical classification and pseudocoloring. The mineral abundance maps highlight areas most likely to contain minerals of interest. Processing was performed for the whole region of interest, and for its part in the central scene alone, which covers about 85 percent of the concession areas. In both cases, statistics for PCA were conducted on a subset of the data, minimizing extraneous factors such as large drainages and urban sprawl, and then applied to the whole region. Mineral abundance maps for the area have been built into a geographic information system (GIS), along with other remotely sensed data, public-domain regional geophysics, geologic, and infrastructure data, scouted geochemistry samples, and a leveled and continuous SRTM digital elevation model. Mapping for the occurrence of mafic-ultramafic rocks was achieved by a combination of PCs 1, 4 and 2 of ASTER bands 2, 4, 5, and 8. Clusters of anomalous contents of selected minerals are draped over the digital elevation model and indicate that the northeast-dipping rock units are covered by a laterized sequence constituting the main exploration targets. This targeting exercise revealed a number of favorable sites that are currently under exploration by mining companies.

Abstract The mineralogy of a leach cap over a subeconomic Mo-Cu porphyry deposit in the Grizzly Peak caldera in the central Colorado Rocky Mountains provides evidence of mineralization and indications of natural acid drainage potential. Airborne hyperspectral imaging (HSI) remote sensing is used to construct a spatially complete map of the leach-cap iron and clay minerals within the mineralized, acid-generating alteration zone. The mixtures of jarosite, goethite, and hematite provide direct indications of mineralization and acid source locations. The clay minerals illite (sericite), kaolinite, dickite, and pyrophyllite further characterize the alteration and appear to correlate with acid seeps. The illite chemistry is analyzed by mapping chemical substitutions of iron for aluminum and is highly correlated with acid sources. When integrated with mapped and interpreted structures, the controls on the acid drainage are revealed.

Abstract This work aims to assess geobotanical anomalies related to high-sulfidation gold mineralization in terranes covered by almost virgin tropical rain forests located in the Tapajos gold province, northern Brazil. Landsat-5 TM data were processed using recently developed algorithms tailored to enhance both specific and anomalous vegetation spectral responses. The method consists of a four-stage successive concentration of information and reduction of data dimensionality, as follows: (1) spectral indexes are prepared to enhance responses of the vegetation characteristics, such as pigments and leaf moisture; (2) a principal component analysis (PCA) is applied on the set of indexes; (3) another PCA is applied to the original NIR and SWIR bands (SWIR; TM 4, 5, and 7), thus obtaining a PC1 as a texture-rich albedo image; and (4) a low-pass proportional convolution filter is applied to selected principal component images derived in the second step. The final product yielded through this technique indicates the presence of differentiated vegetation covers. In the study area, field information showed a near-perfect spatial correlation between color classes highlighted in this Landsat image product and hydrothermal alteration facies identified in outcrops. Although the product is unable to provide direct indication of any rock type, minerals, or presence of hydrothermal alteration, it proved suitable to outline areas where geobotanic anomalies associated to mineralogical variations occur. The method and its products are here considered an important aid for exploration planning and follow-up investigations in the field in terranes covered by equatorial forests.

Abstract The ASTER sensor, which has been aboard the TERRA satellite since late 1999, has 15 spectral bands that cover 14 different wavelength regions in the 0.52- to 11.65- μ m range. Because there are only three primary colors that humans can directly observe simultaneously, displays of triplets of spectral parameters as red, green, and blue are commonly used to enhance the appearance of specific minerals of interest, where they are exposed on the ground. To aid this process, we have created 14 brightness codes (one for each different ASTER spectral band) and 46 spectral ratio codes (36 nonreciprocal spectral ratios of nine spectral bands between 0.52–2.43 μ m and 10 nonreciprocal spectral ratios of five spectral bands between 8.125–11.65 μ m), which divide the mineral library spectra into deciles. Each decile of each spectral band or ratio is labeled from 9 for the highest decile, down to 0 for the lowest decile. A triplet combination with codes of 9, 0, 0 can be displayed as red, blue, green (RGB), respectively, which makes the mineral of interest red in the resulting image, with few (usually well less than 10% of the minerals in the library) false positives. Examples of how spectral ratio codes may be applied are demonstrated using ASTER data of the west-central Powder River basin in Wyoming for the enhancement of hematite (in red beds), gypsum, quartz (in sandstone), and calcite (in limestone). Supervised classification derived from training sets identified on spectral ratio images selected on the basis of ratio codes of ASTER data produced a lithologic map of the study area in the west-central Powder River basin that had an accuracy of 63.3 percent, compared with field data. We conclude that the supervised classification provides a more accurate lithologic map than we could have produced by using a traditional geologic map from which to pick training sets.

Abstract Recent exploration efforts at the El Capitan iron deposit in New Mexico have identified potentially economic Au, Ag, and Pt values in field samples and drill core. The El Capitan deposit is characterized by iron-rich calcsilicate skarn mineralogy and a later stage hematite-calcite dominant assemblage. Gold-platinum mineralization occurs mainly in the hematite-calcite assemblage. Although this assemblage occurs in skarn as apparently primary skarn minerals, it is believed to largely represent a later stage fracture-filling mineralizing event. An airborne hyperspectral survey was contracted to fly over the project area to aid in the exploration effort. The area is fairly rugged and contains a significant amount of vegetation so that conventional mineral exploration methods are tedious and time consuming. Six overlapping flight lines of airborne hyperspectral data were acquired to assist in the exploration effort. To help characterize important mineral absorption features, reflectance spectra of various rock and mineral samples were collected using a field spectrometer. Analysis of these field spectra has helped in the interpretation of the airborne data. The focus of the hyperspectral mineral mapping was on carbonate rocks, calc-silicate skarn minerals, hematite, and goethite. The analysis of hyperspectral data over the existing workings has helped direct analysis for exploration throughout the district. Several anomalous areas of hematite and/or goethite, and calc-silicates were identified in the data. The anomalies were field checked and several were found to have anomalous gold and platinum values. The focus of future exploration efforts, including drilling, will be on some of these specific targets identified in the hyperspectral data.

Abstract Multibeam RADARSAT-1 and multispectral Landsat TM-7 imagery have been used to map surficial materials in the Schultz Lake area of Nunavut (1:250,000 NTS map sheet 66A). Representative training areas of distinctive surficial materials (bedrock, boulder fields, organic deposits, sand-gravel, thick and thin till) have been identified through the analysis of aerial photographs, satellite images, and field mapping information. These training areas have been used to perform a maximum likelihood classification using combined multibeam radar and multispectral satellite imagery to produce a predictive map of surficial materials. The application of this methodology has resulted in a classified predictive map of the surficial materials in the area. Based on the training areas, the overall accuracy of the predictive map is greater than 80 percent. Classes with the least accuracy included sand-gravel and thin till due to their spectral and textural similarities with thick till. This remote predictive mapping approach can be used for ground follow-up in surficial mapping and mineral exploration programs.

Abstract This work focuses on the analysis of Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) images collected over salt lakes occurring in the Sud Lipez region (southwestern Bolivia). Thanks to its extreme aridity, lack of vegetation cover, high latitude (>4,000 m) and wide variety of brine geochemistry, this region proves to be an ideal place to demonstrate the capabilities of ASTER visible, near, and short-waved infrared bands in mapping mineral occurrences. ASTER images obtained over two major salt lakes, namely Capina and Pastos Grandes, have been corrected and properly calibrated using field spectrometry. Their further processing using pixel purity indices reveals end-member spectra that closely match those of mineral species—such as gypsum, calcite, and ulexite—known to occur in this context. The classification of the multispectral scenes on the basis of these selected end members reveals that a detailed mineral mapping in such evaporitic environments can be achieved with high confidence and good spatial accuracy. When compared to the very schematic geologic maps recently made available from field work, the remote sensing data indicate a very good correlation while revealing instructive patterns of facies variations that can serve as a basis for further field exploration work. The potential of the Bolivian section of the Andes cordillera for borate deposits is high but still poorly explored. Borate concentrations often form discontinuous pods and lenses within sediments but can locally evolve into continuous decimetric strata. Though most of the time such deposits are covered by a burden of sedimentary and evaporitic materials, it is believed that multispectral remote sensing can be extremely useful in spotting limited outcrops of boron minerals.

Abstract This study demonstrates the application of airborne hyperspectral data for the generation of accurate remote predictive geologic maps, which can be used to assist regional mapping by (1) giving detailed spatial and spectral information, (2) focusing future mapping projects, and (3) highlighting areas of economic potential. The study area is located in southern Baffin Island and comprises a diverse assemblage of lithologic units that are part of the northeastern segment of the Paleoproterozoic Trans-Hudson orogen. Two steps were required to generate remote predictive geologic maps from the hyperspectral image: the extraction of image end members and the application of spectral mixture analysis to generate fractional abundance maps; and converting the fractional abundance maps into predictive geologic maps. Eleven geologic units were extracted from the image data as end members. These end members were identified based on characteristic spectral features and comparisons with field and laboratory spectra. The predictive map correlates well with the existing published map, but more extensive exposures of potentially economic peridotite and carbonate units were found. Lichen-rock mixtures were used to map quartzofeldspathic units that are covered by thick lichen coatings in this region.

Abstract Plagioclase feldspar composition is a potentially useful geobarometer and indicator for metasomatism in many styles of gold and base metal deposits. However, routine measurement of feldspar composition for exploration purposes is challenging using conventional laboratory methods such as X-ray diffraction, because of the time and cost to conduct these analyses. The thermal infrared (TIR) wavelength region between 7 and 14 μm provides diagnostic spectral information about silicates, including the composition of feldspars. To capture this information for routine logging of diamond drill core, CSIRO developed a line-profiling, bidirectional reflectance TIR scientific prototype called the TIRLogger. This paper examines the efficacy of the TIRLogger’s bidirectional reflectance measurements and the derived mineral information products through comparison with laboratory MicroFTIR™ emissivity measurements, X-ray diffraction, X-ray fluorescence, and electron microprobe data of the same samples. The results show that the plagioclase XRD 111 and 131 hkl peaks are significantly correlated with the normalized height of the background-corrected TIR reflectance peaks at 9.6 and 9.9 μm, with increasing Ca content in plagioclase associated with a relatively lower 9.6 μm peak. TIRLogger analysis of two diamond drill holes intersecting Archean greenstones from the St. Ives area in Western Australia show that Ca plagioclase, together with dolomite and quartz, are developed in areas of anomalously high Au content, whereas albite and calcite were detected in unaltered rocks, except for ultramafic rocks, which also shows that dolomite is developed.