Jump to content

Course:EOSC311/2026/Hyperspectral Imaging

From UBC Wiki

Introduction

Infrared absorbance spectrum of a soil sample with Montmorillonite (clay mineral). The peaks act as a chemical fingerprint, enabling scientists to identify the mineral composition of the sample through spectroscopic analysis.

Mineral mapping is the identification, analysis and visual depiction of mineral distribution. [1] Traditionally, this process occurred using manual processes such as the physical sampling of rocks and other lab techniques like scanning electron microscopy. However, these traditional techniques can be costly, time consuming and risky for geologists. [1] In addition to this, they often necessitate the destruction of rocks in order to extract the samples needed for analysis.

To address these limitations, geologists today employ analytical chemistry and spectroscopic analysis, chemical fields that aid in determining the qualitative and quantitative components of various substances. Spectroscopy, specifically, examines the interactions between light and matter, such as the absorption or emission of electromagnetic radiation. [2] In traditional laboratory spectrometers, atoms or molecules are irradiated with light, causing the absorption, emission, transmittance or reflection of radiation. [3] The spectroscopic detector then measures these light intensities across wavelengths, converting them into electrical signals which are visualized as spectra that scientists can analyze. [3]

In minerals, which have rigid crystalline structures, shining a light on them leads to the same interactions. [2] Because minerals absorb and reflect light across different wavelengths, every mineral has a unique chemical fingerprint (spectral signature), allowing them to be identified by the wavelength at which they reflect light. [2] One method of leveraging spectroscopy is Hyperspectral Imaging (HSI), which is a technology that was introduced in the late 1970s. By combining spectroscopic analysis with digital imaging, this technology is able to utilize chemical fingerprints to remotely and noninvasively determine the identity of minerals. [4] Unlike traditional methods, HSI increases the accuracy and efficiency of mineral exploration by eliminating the need for time consuming, manual exploration.

Statement of Connection

Though chemistry and geology may appear to be very different fields, both disciplines often rely on each other to explain complex processes and entrench theories. Spectroscopic analysis is often used in chemistry as a standard method for determining the identity, quantity and structure of chemicals, however it also has significant applications in geology. These applications aid in revealing quantitative and qualitative characteristics of minerals, such as mineral identity and elemental concentration. When applied to mining, this allows geologists to safely, efficiently and remotely complete the exploration stage of mining.

The connection between spectroscopic analysis and geology was drawn because, as chemistry majors, we have a lot of experience conducting spectroscopic analyses. Because of our experience, we were able to recognize that practical applications of this chemical process exist outside of the chemistry lab. We chose hyperspectral imaging as the method of evaluating minerals because it perfectly utilizes the chemical principles of spectroscopic analysis to aid geologists, providing them with greater efficiency and accuracy than traditional geological methods such as physical evaluation of rocks.

Geological formation of mineral deposits

Today we depend on different metals to support our livelihood, yet the geological processes needed to concentrate these metals into economically viable deposits are rare. [5] While elements like copper are widespread throughout the Earth's crust, they are mostly distributed in low concentrations within minerals. [6] Minerals are naturally occurring inorganic solids with a specific chemical composition and crystalline structure. [7] These minerals are typically spread out within ordinary rocks in trace amounts, making the extraction of metals disadvantageous in most locations. Mining is only favourable when geological processes such as plate tectonics, magmatism, and hydrothermal circulation act over millions of years to concentrate these scattered elements into localized mineral deposits. [8] By learning how copper and iron ore bodies are formed, we can better understand how important Earth's dynamic systems are in supplying resources that fuel many industries across the world.

Convergent plate boundary exhibiting subduction process. The blue dots are the release of volatile compounds, while the red dots are the magma chambers formed from the lowering of the melting point of the mantle rock.

The formation of economically viable copper deposits is initially driven by plate tectonics. At a convergent plate boundary, where a dense oceanic plate collides with a thicker and more buoyant continental plate, the oceanic plate is forced downward into the mantle in a process called subduction. [9] The oceanic crust is exposed to accelerating lithostatic pressure and heat, conditions that force volatile compounds (primarily water, sulfur and carbon dioxide) out of the sinking slab and into the overlying mantle wedge. [10] The introduction of such compounds lowers the melting point of the mantle rock, allowing for partial melting and the formation of molten rock. [10] This magma rises through the crust and forms chambers several kilometres below the surface as a porphyry intrusion. [11]

As the magma rises and evolves, copper can become concentrated within the melt and later within the hydrothermal fluids released from it. Copper is a chalcophile element, meaning it has a strong affinity for sulfur, [12] so it’s commonly transported in sulfur rich magmatic and hydrothermal systems rather than being incorporated into common rock forming minerals such as quartz and feldspar. As the magma cools and crystallizes, these common minerals form first, leaving the remaining melt enriched in water, volatiles, sulfur, and dissolved metals such as copper. [11]

Circulation of hot fluids through the magma chamber itself and the surrounding rocks. Scavenging copper, molybdenum, and gold primarily from the magma and lead, zinc, and silver from the surrounding rocks, creating a metal rich brine.

As the magma body cools and crystallises, it releases the trapped and superheated mineral rich water and gases, increasing the pressure within the magma chamber until the surrounding rocks fractures.[11] These hot fluids can now escape through the fractured crust, creating a circulating hydrothermal system capable of transporting high concentrations of dissolved metals like copper. [13] They leach trace metals out of the surrounding host rock since it possesses characteristics of an aggressive chemical solvent, binding them into highly mobile complexes. Driven upward and outwards, these metal rich fluids experience a sudden drop in pressure, temperature, or come into contact with surrounding rocks that changes its chemistry (reacts), causing the solution to destabilise. The metals precipitate out and form solid minerals sequentially, with copper exiting first (zonation). [13] Over time, these minerals fill fractures to form networks of veins or spread through the host rock, creating concentrated ore regions. This process creates the world’s largest copper sources, like porphyry copper deposits where copper is tied up inside minerals like chalcopyrite (CuFeS2) and bornite (Cu5FeS4). [11] Iron can also be transported and concentrated by hydrothermal fluids, but the world's largest iron deposits formed through different sedimentary processes. [14]

Banded Iron Formation: sedimentary rock composed of alternating bands of dark, iron rich minerals (magnetite and hematite) and lighter coloured, silica rich layers (chert).

While copper accumulation depends on magmatic and hydrothermal processes, Earth’s largest iron deposit forms through a different set of events: ancient ocean chemistry, sedimentation, and later tectonic changes. Much of the world’s mined iron comes from the Banded Iron Formations formed billions of years ago.[14] During the Archean eon the oceans lacked free oxygen, allowing for large amounts of chemically reduced, soluble ferrous iron (Fe2+) to dissolve into the seawater from deep sea hydrothermal vents. Right at the start of the Paleoproterozoic eon, Earth undergoes the great oxidation event through the appearance of photosynthetic cyanobacteria, introducing free oxygen into the marine environment. This dissolved oxygen reacted with the soluble iron, transforming it into insoluble ferric iron (Fe3+). The oxidized iron precipitated out of the water column in large, continuous blankets of magnetite (Fe3O4) and hematite (Fe2O3), alternating with layers of silica rich chert. Over billions of years, burial, tectonic activity, and metamorphism compressed and converted these sediments into the iron ore deposits that supply a lot of the world's iron today. [15] [16]

Principles of Hyperspectral Imaging

This spectrum shows the absorbance of elemental iron between 400 and 700 nm of wavelength.

Hyperspectral imaging refers to a technique that analyzes materials through the combined use of spectroscopy and cameras. At its core, spectroscopy studies the effects of light on the material of interest. When light, otherwise known as electromagnetic radiation, shines on physical matter, it can be absorbed or emitted, and scientists analyze this response. Recall that light varies in wavelength, or inversely, frequency. Scientists adjust the wavelength of light to obtain a different signal for each wavelength. This spectrum of response is known as a spectral signature, and it is unique to each material. Essentially, it functions as a fingerprint for identifying the composition of the material being analyzed. When used on an unknown substance, the spectrum produced can be compared to the patterns and characteristics of known spectral signatures to determine its identity. [17]

An example of a three-dimensional image created by hyperspectral imaging.

HSI combines spectroscopy with digital cameras, enabling the technology to capture mineral identity and location. This aids geologists in creating a 3D map of an area, with spectrographic data represented in each pixel. [1] Because of its ability to aid in the detection of material, HSI is employed not only in geology, but in a wide range of fields. Applications include, but are not limited to, quality control in food products, vegetation health in agriculture, regulations of water preservation, and analyses in medical diagnoses. [17]

Applications of Hyperspectral Imaging

Satellite Hyperspectral Imaging

Satellite imaging, one of the mediums of HSI, entails the use of a space born probe to determine the spectral signatures of minerals and other natural materials. [18] Satellite hyperspectral imagers primarily employ diffractive gratings or reflective prisms as dispersive elements, which are the methods by which light from an image is separated into its wavelengths or “spectral channels”. [18]

Image depicting satellite used for hyperspectral imagery.

In satellite imaging, scanning of areas of interest typically occurs using push-broom scanning. [18] With this scanning technique, the satellite camera collects image data of the area of interest line by line rather than all at once. [19] Namely, spatial lines are imaged one at a time and are subsequently divided into spectral components. Once the components reach the sensor array, 1-directional spatial imaging is carried out, creating part of a 2-dimensional image. [19] To achieve 2-dimensionality, the satellite camera scans the desired area along an orbital path. This results in high precision spectral imaging and therefore, high spectral and spatial resolution due to the slow orbit of the satellite allowing more photons to hit detectors. This longer dwell time increases the signal to noise ratio, leading to high spatial and spectral resolution. [4]

This form of imaging is advantageous for visualizing large areas of land (about 1000 km2) and is typically employed as the initial stage of mineral detection. [4] Some uses of these satellite systems in mining are for identification of hazardous zones. [4] Identification entails the ability to diagnose the presence of environmentally and biologically harmful minerals in lakes and mines. This also includes the ability to map the different chemical components of waste. Some other applications are found in pH and mineralogical trend mapping.

Aerial Hyperspectral Imaging

Though satellite HSI offers wide coverage of areas of interest, aerial HSI, also known as unmanned aerial vehicles (UAVs), bridges the gap between space operated systems and ground systems, allowing geoscientific experts to obtain a closer look into mines. [20] As opposed to satellite HSI where the sensors are located in space, aerial HSI sensors are placed on drones at significantly lower altitudes. This allows hyperspectral sensors to capture large areas of land with increased specificity. [20] At the same time, aerial HSI does not limit sampling of metals to one small area, as is done with drill core logging, but allows the remote analysis of a desired spatial distribution whilst delivering higher spatial and spectral resolution. By employing aerial HSI, geoscientific experts have measured control over the speed, altitude and route of the drones. [21] In doing this, they can elude atmospheric obstacles, such as clouds, that typically distort satellite HSI signals. [21] 

With respect to mining, aerial HSI is primarily used in open pit mines for 3-D mapping of walls. [22]This function is highly significant as open pit mine walls are typically steep, and are therefore likely to be hazardous to geologists aiming to obtain rock samples. After capturing spectral data and creating digital outcrop models, HSI enables geologists to create hyperclouds, which are 3-D models improved by hyperspectral reflectance data. [22] Mapping with UAVs allows for the remote analysis of mineral concentrations and identities, whilst the use of hyperclouds allows for more precise differentiation of ore from rocks and provides the ability to connect spectral signatures to drillhole data, leading to better mineral extraction. [22]

Drill Core Logging

Close up image depicting three types of hollow cylindrical coring drill bits used for extraction of rock samples.

Drill core logging is the most precise application of HSI as it involves the physical extraction of rock samples. In this process, a hollow cylindrical coring drill is used to collect a cylindrical rock sample that can extend for thousands of feet. [23] This sample allows geologists to determine the composition of rocks beneath the earth's surface. In traditional core logging, drill cores are analyzed manually in processes that are time consuming and subjective because they require geologists to perform manual inspections. [1] This causes traditional processes to be expensive and prone to human error.

Conversely, the application of HSI to drill core logging automates the process by handing off drill cores to hyperspectral sensors rather than geologists. In this process, drill cores are scanned by hyperspectral cameras located at ground level. This minimizes the distance between the sensor and the core, thus achieving a high signal to noise ratio. [23] HSI analyzes the core line by line, helping the system to determine mineral identities, compositions and structures. Though not remote, drill core logging enables geologists to noninvasively analyze minerals in drill cores in a matter of seconds. [23]

Integration

Though separate, these three applications of HSI can function together to maximize efficiency and decrease risk during mining exploration. These techniques also eliminate the need for random exploration to determine mineral location and composition, therefore reducing costs and increasing environmental viability. The integration of the three techniques would entail beginning the process of mineral mapping with satellite HSI. Starting in space, satellites scan over large areas of land to identify areas of interest. UAVs are then used over the areas of interest to determine precise mineral locations. Drill cores can then be obtained using coring drills and HSI can scan samples to analyze and determine mineral composition, identity, and concentrations.

Real examples of HSI usage

With the development of HSI over the past few decades, it has become an advanced tool in a wide variety of fields, such as mining, agriculture, medicine, food and many more. This section focuses on the applications in mining, primarily for the rapid, yet thorough, spectroscopy of rock samples.

An aerial view of the Highland Valley Mine, located in the southern portion of British Columbia, Canada.

The Highland Valley mine, located in southern British Columbia, is perhaps the closest example to the University of British Columbia. It is well known for its copper deposit, being one of the largest in all of Canada, but it is also home to a variety of minerals. Using HSI in the short-wavelength infrared region, geologists were able to identify white mica, kaolinite, and quartz. Identifying areas of certain materials, like prehnite, can be indicative of mineralization on a larger scale. Extrapolation of this data can pinpoint materials on a map, guiding the directions of the mining exploration.[24]

An image of Olympic Dam Mine, located in southern Australia.

Another example of the mining industry shifting towards using HSI is with Rio Tinto, a prominent mining company. In 2022, it was announced that Rio Tinto would be collaborating with Pixxel, a satellite company. As opposed to regular, multispectral techniques, Pixxel uses high-resolution hyperspectral imaging aboard their satellites to observe the Earth. This results in an impressive resolution of five meters, which is fifty times more detailed than previous technologies. These new developments are designed to aid Rio Tinto in assessing areas of minerals around the world, critically influencing the next generation of mine discoveries.[25]

Hyperspectral imaging has also been used in Oceania, with a recent application being at the Olympic Dam mine in southern Australia. This location is known for its iron oxide and copper-gold deposits. Specifically, drill cores containing copper were obtained and analyzed by HSI. Interestingly, the geologists used the data from HSI to teach machine learning models to accurately predict the concentration of copper when analyzing drill cores. They found this combination of algorithms and HSI to be a promising technique in acquiring vast amounts of data about the rocks within a short timeframe.[26]

Conclusion

Hyperspectral imaging is becoming an increasingly important and established tool in mineral exploration globally. By capturing sunlight or artificial light reflecting off the exposed mineral surfaces, this technology records a continuous sequence of adjacent wavelengths across the electromagnetic spectrum to instantly transform raw rock faces into dynamic and compositionally precise maps. [27] Developments in HSI sensors have provided higher spatial and spectral resolution to mining geologists and engineers, supporting them in the identification of smaller geological features such as veins and gossans and differentiation of minerals in more detail.[28][29] Traditional mineral mapping methods possess disadvantages that the HSI process can overcome! Systematic geochemical sampling or manual field mapping is an example of an older mineral mapping method that required many individuals physically present collecting soil and rock chip samples from different grid intervals to produce a map.[30] This method is costly, time consuming and sometimes destructive to the land, while HSI offers a non-invasive, cost effective and faster alternative that changes how geologists analyse the terrain. Another example used for mineral mapping was the ASTER satellite imagery, only capturing limited discrete broad bands in the shortwave infrared spectrum and possessing lower resilience, as its shortwave infrared detectors  stopped functioning in 2009 due to anomalously high temperatures.[31] Modern HSI overcomes these limitations through providing an ever evolving robust system with detector arrays that deliver uninterrupted, high resolution and continuous spectral mapping.[29] Its use in satellite imaging, aerial drone surveys, and drill core logging shows how this technology can be applied at different stages of exploration, from scanning large regions to studying specific rock samples in detail. [19][20][23] As demand for metals like copper and iron grows, HSI will support more efficient, cost effective, environmentally responsible and versatile exploration. Through this assignment we were able to understand this topic through connecting both physical geology and quantitative chemistry principles, showing how different fields of study aid in the overall growth in industries and people's livelihood.

References

  1. ↑ 1.0 1.1 1.2 1.3 Macheyeki, Athanas S.; Li, Xiaohui; Kafumu, Dalaly P.; Yuan, Feng (2020). "Chapter 3 - Conventional and nonconventional exploration techniques–principles". Applied Geochemistry: Advances in Mineral Exploration Techniques: 87–149.
  2. ↑ 2.0 2.1 2.2 Madejova, J.; Gates, W.P.; Petit, S. (2017). "Chapter 5 - IR Spectra of Clay Minerals". Developments in Clay Science. 8: 107–149 – via Elsevier Science Direct.
  3. ↑ 3.0 3.1 "How Spectroscopy Is Revolutionizing Modern Research". OceanOptics. June 17, 2026. Retrieved June 17, 2026.
  4. ↑ 4.0 4.1 4.2 4.3 Koerting, Friederike (2024). "VNIR-SWIR Imaging Spectroscopy for Mining: Insights for Hyperspectral Drone Applications". Mining. 4: 1013–1057.CS1 maint: display-authors (link)
  5. ↑ Cullers, Robert L. (2024). "Economic Geology". EBSCO. Retrieved 12/06/2026. Check date values in: |access-date= (help)
  6. ↑ "12.2 Minerals". Lumen Learning. Retrieved 11/6/2026. Check date values in: |access-date= (help)
  7. ↑ "What is the difference between a rock and a mineral?". 7/11/2024. Retrieved 12/6/2026. |first= missing |last= (help); Check date values in: |access-date=, |date= (help)
  8. ↑ Neser, Laura (13/12/2022). "Energy and Mineral Resources". Pressbooks at Virginia Tech. Retrieved 11/6/2026. Check date values in: |access-date=, |date= (help)
  9. ↑ Panchuk, Karla (2021). "4.4 Plates, Plate Motions, and Plate-Boundary Processes". British Columbia/Yukon Pressbooks. Retrieved 12/6/2026. Check date values in: |access-date= (help)
  10. ↑ 10.0 10.1 Zellmer, Georg F.; Edmonds, Marie; Straub, Susanne M. (1/1/2015). "Volatiles in subduction zone magmatism". Geological Society, London, Special Publications. 410. Check date values in: |date= (help)
  11. ↑ 11.0 11.1 11.2 11.3 Torvela, Taija (20/10/2023). "Porphyry copper deposits". Youtube. Retrieved 12/6/2026. Check date values in: |access-date=, |date= (help)
  12. ↑ Wieser, Penny E.; Jenner, Frances E. (2/12/2020). "Chalcophile Elements: Systematics and Relevance". Encyclopedia of Geology (second edition): 67–80. Check date values in: |date= (help)
  13. ↑ 13.0 13.1 Eagle Plains Resources (20/9/2021). "Porphyry and Epithermal mineral Deposits". Youtube. Retrieved 12/6/2026. Check date values in: |access-date=, |date= (help)
  14. ↑ 14.0 14.1 Ho, J. D. (2024). "Iron deposits". EBSCO. Retrieved 12/6/2026. Check date values in: |access-date= (help)
  15. ↑ Earth Science Classroom (5/10/2023). "Banded Iron Formations (BIFs)". Youtube. Retrieved 12/6/2026. Check date values in: |access-date=, |date= (help)
  16. ↑ Earle, Steven (1/9/2015). "20.1 Metal Deposits". Open Text BC. Retrieved 12/6/2026. Check date values in: |access-date=, |date= (help)
  17. ↑ 17.0 17.1 Bhargava, Anuja; Sachdeva, Ashish; Alsharif, Mohammed H.; Uthansakul, Peerapong; Uthansakul, Monthippa (June 2024). "Hyperspectral imaging and its applications: A review". Heliyon. 10 – via Elsevier Science Direct.
  18. ↑ 18.0 18.1 18.2 Krupnik, Diana; Khan, Shuhab (2019). "Close-range, ground-based hyperspectral imaging for mining applications at various scales: Review and case studies". Earth-Science Reviews. 198 – via Elsevier Science Direct.
  19. ↑ 19.0 19.1 19.2 "HySpex Push-Broom (Linescan) Hyperspectral Cameras". SphereOptics. Retrieved June 17, 2026.
  20. ↑ 20.0 20.1 20.2 Park, Sebeom; Choi, Yosoon (2020). "Applications of Unmanned Aerial Vehicles in Mining from Exploration to Reclamation: A Review". Minerals. 10: 663 – via MDPI.
  21. ↑ 21.0 21.1 Brass, Kristaps (February 19, 2025). "Hyperspectral Cameras and Drones: A Practical Guide". Sphengineering. Retrieved June 14, 2026.
  22. ↑ 22.0 22.1 22.2 "Mining and Mineral Exploration with Hyperspectral Imaging Solutions". Specim. June 14, 2026.
  23. ↑ 23.0 23.1 23.2 23.3 Matimbi, Dzhavhelo K.; Carranza, Emmanuel John M. (2026). "Integration of drill-core hyperspectral and geochemical data by deep learning to enhance drill-core mineral mapping". Applied Computing and Geosciences. 31 – via Elsevier Science Direct.
  24. ↑ Lypaczewski, Philip (April 2020). "Characterization of Mineralogy in the Highland Valley Porphyry Cu District Using Hyperspectral Imaging, and Potential Applications". Minerals – via MDPI.
  25. ↑ "Rio Tinto partners with Pixxel to investigate hyperspectral satellite tech capabilities". IM International Mining. Retrieved 2026-06-10.
  26. ↑ Prado, Elias M. G.; Filho, Carlos R. S.; Carranza, Emmanuel J. M. (December 2023). "Ore-Grade Estimation from Hyperspectral Data Using Convolutional Neural Networks: A Case Study at the Olympic Dam Iron Oxide Copper-Gold Deposit, Australia". Economic Geology: 1899–1921 – via Geoscience World.
  27. ↑ Earth Mapping Resources Initiative (9/12/2024). "Hyperspectral Mapping". USGS. Retrieved 14/6/2026. Check date values in: |access-date=, |date= (help)
  28. ↑ Laakso, Kati S. (1/1/2015). "Applying Hyperspectral Remote Sensing Techniques for the Detection of Hydrothermal Alteration Zones and Gossans in Northern Canada". University of Alberta Library. Retrieved 14/6/2026. Check date values in: |access-date=, |date= (help)
  29. ↑ 29.0 29.1 "Mining and Mineral Exploration with Hyperspectral Imaging Solutions". Specim. Retrieved 14/6/2026. Check date values in: |access-date= (help)
  30. ↑ Goss, Brian (12/7/2025). "The Role of Geochemical Soil Sampling in Discovering Ore Deposits". Rangefront Mining Services. Retrieved 14/6/2026. Check date values in: |access-date=, |date= (help)
  31. ↑ "ASTER User Advisory". Earth data, NASA. 14/1/2009. Retrieved 14/6/2026. Check date values in: |access-date=, |date= (help)


This Earth Science resource was created by Course:EOSC311.
Retrieved from "https://wiki.ubc.ca/index.php?title=Course:EOSC311/2026/Hyperspectral_Imaging&oldid=898984"