Economic minerals remain the necessary and relevant resource of the industry and the entire modern civilization. Mining and processing of minerals imply a fundamental aspect of social production, and the resource base is fuel for transport and social infrastructures. The globalization of the world economy estimates transnational cooperation in the search and exploration of new deposits in the most promising regions of the planet, including the mineral resources of the World Ocean. Remote sensing is a non-contact study of the Earth, its surface and subsoil, individual objects and phenomena, conducted by registering and analyzing their own or reflected electromagnetic radiation. The application of the industrial minerals should be scrutinized to detect its effectiveness for geology and economics regarding the highly prospective character of the remote sensing methods.
How Minerals are Identified With Relationship to Remote Sensing
Luster identifies both the internal structure and the nature of the reflective surface of the mineral. Minerals with a metallic luster are easily distinguished, while with metallic and metal-like luster often have a black or dark line (magnetite, galena, graphite). Minerals with a white and color line usually have a non-metallic luster (gypsum, sulfur, cinnabar). In the group of minerals with a metallic luster, the exceptions list includes native gold, copper, silver, platinum, chalcopyrite, and faded ore. Having a metallic luster, they give a color line: gold – greenish, silver – silver-white, copper – copper-red, chalcopyrite – greenish, faded mineral – dark brown. Non-metallic luster is divided into: polymetallic (the mineral has a luster of metal, but it has colored powder and line), diamond, glass, oily, silky, nacreous, and matted.
In turn, remote sensing or non-contact study of the Earth, its surface and subsoil, individual objects and phenomena is carried out by recording and analyzing their reflected electromagnetic radiation. There is recorded either natural radiation, determined by the natural illumination of the Earth’s surface by the Sun, thermal radiation – the Earth’s “personal” radiation, or artificial radiation, which is created when an area is irradiated with a source located on a recording device carrier (Gupta, 2017). Registration can be performed using technical means installed on aircraft and spacecraft, or the Earth’s surface as an alternative. The image can be represented in the form of a two-dimensional analog recording, for example, photographic or digital record on magnetic storage devices.
Its components include the source of electromagnetic radiation, the process of propagation of radiation and its interaction with the substance of the object, the response signal, data recording and providing to consumers. In this model, the source generates electromagnetic radiation with a high level of energy over the entire wavelength range, while the radiation intensity embodies a known value that does not depend on the wavelength (Landgrebe, 2005). The radiation does not “communicate” with the atmosphere and propagates through it without energy loss. The incident radiation interacts with the substance of the object, as a result of which the reflected or its secondary radiation appears, being homogeneous over the entire wavelength range. The radiation falls on the sensor right from the object; the sensor records the spatial information during the process (Sabins, 2007). The ideal sensor should have a simple and compact design and have high accuracy. The recorded data are transmitted to the ground station, where it is instantly transformed into an interpretable form. It allows identifying all parts of the object being studied by the physical, chemical, and biological properties. The data is provided to consumers in this form.
The crystal structure has a direct relation to its coloring. “Aggregative” coloring depends on the structural features of the mineral aggregate. It appears as a result of light absorption by an extraneous mineral phase or as a result of scattering, refraction, diffraction, and interference of light at the boundaries of mineral phases and inclusions (the so-called “false colors”) (Sabins, 1999). In turn, “own” colors are characterized by selective absorption of light by the mineral or light filter (“spectral window” for transmitted light). The spectrum of light absorption determines the character of the spectral window. The Butterfly Effect is of particular interest in this regard; it is represented by the light effects that change the visible color of minerals. The light effects are the result of changes at the boundaries between the crystals, cleavage cracks, and surface oxidation films. For example, the color of milky white quartz is the result of light scattering on the gas-liquid inclusions; the color of red halite is produced by the embodiments of hematite (absorption by mineral impurities).
As for the mineral groups identifying, the next few points should be noted. There are solid, liquid, and gaseous groups of mineral formations circulating in nature. Solid minerals can be crystalline and amorphous. Crystalline consist of many identical structural elements, forming an ordered spatial (crystal) lattice. There are atomic, ionic, and molecular types of lattices, which determine the anisotropy (different properties) and isotropy (identical properties) of crystals.
Meanwhile, any natural and artificial crystals possess the form of polyhedra. They can be isotropic and anisotropic. Amorphous minerals are always isotropic. The ability of substances with the same chemical composition to crystallize in different forms is called polymorphism (multiformity). The examples include diamond and graphite, pyrite and marcasite, calcite and aragonite.
The different structure of polymorphic species explains their multiple properties. Some substances of varying chemical composition can form similar crystallographic forms. Such materials can create mixed patterns containing the primary components in different proportions. This phenomenon is called isomorphism, and mixtures are called isomorphic. For instance, it could be feldspars, the isomorphic series of which are formed when the albite and anorthite molecules are mixed. In natural conditions, there often grow not quite regular crystalline forms, which have some defects; but with any flaws, the angles between the corresponding crystal faces of the same substance remain the same and constant. This law of constancy of face angles makes it possible to establish the ideal form of crystals and accurately diagnose the smallest mineral grains.
Spectral analysis assists in the identification of the minerals via referring to their ability to absorb and reflect light. The principal basis for the spectral analysis of data is the assumption of a one-to-one correspondence between the reflected signal (in general, intensity and polarization) and the composition of the reflecting surface. There will be observed different results for distinct reflective materials regarding a sufficiently large number of spectral channels, which measure the intensity of the reflected signal. Following this, hyperspectral analysis means such spectral analysis which is conducted for the number of spectral channels (the number of wavelengths for which the intensity is measured) from a few hundred to thousands; hyper-spectrometer is a device that simultaneously measures the strength of signal emission for the wavelengths and spatial coordinates of the investigated surface.
The technology for processing of the recorded multispectral images is traditionally based on methods for classifying underlying surface types based on the analysis of the ratios of the brightness in various channels. A distinctive feature of hyperspectral data is the narrow width of the spectral bands and a large number of recorded channels. Based on this, numerous approaches have been developed that implement the analysis of the fine structure of the images pixels spectrum and their classification by comparison with the reference spectral curves (the spectral curve characterizes the relationship between the values of reflection coefficients and the wavelength).
Naturally, the intensity of the reflected signal depends on the magnitude of the backlight signal, where natural light sources like the Sun, the Moon, stars, and artificial lighting can be used. It is convenient to work with natural lighting when conducting full-scale measurements, whereas the use of artificial light is suitable for creating a spectral library of samples and leading of the laboratory studies. The spectral libraries are essential due to the next reasons. Spectral analysis is based on a comparison of the histograms obtained during the survey with the histograms of already known materials. A necessary analysis tool is a spectral library – a database containing information about the reflectivity of materials at different wavelengths. The largest public libraries include the Johns Hopkins University and JPL Spectral Libraries.
Since it is difficult to carry out field measurements under the same conditions, it is necessary to calibrate the data obtained taking into account the intensity and spectrum of the incident light. The central values to be changed during spectral analysis are the wavelength, the energy of the reflected signal, and the spatial coordinate of the surface under study. Therefore, the central object of the hyperspectral analysis is the hypercube – an array of data formed by the intensity values of the reflected signal from a two-dimensional surface, divided into pixels. Each pixel corresponds to the spectral coordinate – the intensity and discrete polarizability. The main feature of hyperspectral analysis, which distinguishes it from multi-zone or multispectral (using only a few spectral channels), is the possibility of using its differential characteristics for image identification in contrast to a multi-zone image, in which they are averaged over wavelength (Villiger, 2008). It leads to a fundamental improvement in the informativeness of the data since not the few (usually not more than ten) of the most informative or straightforward parts of the spectrum are used for recording, but the whole range of the recorded wavelengths.
Regarding the example study, one of the promising areas of application of hyperspectral analysis is remote sensing of the Earth. Shooting is performed from the aircraft, helicopter, satellite, or drone, and the Sun is used as a source of illumination. In remote sensing, it is necessary to take into account the distortion of the signal received by the atmosphere, different for various weather conditions and angles of light incidence. Therefore, close attention is usually given to the algorithms that implement them, allowing to take into account these effects. Regarding geology, the method helps in the diagnosis of rocks and the identification of minerals; hyperspectral data may determine, for example, dense, partially dolomitic, or highly dolomitic limestone.
Differences between Field Studies vs. Air Studies
The insufficiency of solar illumination characterizes the sensing in the field conditions. The same is about the laboratory conditions, where the lighting is weaker than the solar one used in the air studies. If the hyper-spectrometer is designed to work in laboratory conditions, it can also be equipped with a light source. Therefore, the main difference between the air and field studies is in the cause of light that demands its consideration in the construction of the sensing tool (Khan and Jacobson, 2008). The filter system, located behind the input lens, highlights one or several spectral ranges in the incident light, following the tasks assigned to the instrument, and also allows each spectral range to pass through a separate optical system, increasing the measurement accuracy. There is a lens system behind the filter system that connects the adjacent image areas for each filter and creates a single image. Electron-optical amplifier, located in front of the receiver, allows amplifying the signal. It is often necessary when field studies have weaker light sources compared to the air studies when natural lighters perfectly fit for getting of the precise data.
The urgency of creating new types of aeronautical hyperspectral modules is determined by their ability to extract maximum information from the optical radiation rising from remotely sensed objects. Consequently, regarding remote sensing, the development and application of drones technology are characterized by maximum prospects. The drone can perform completely free flights lasting a few hours with the possibility of increasing this figure (Jordan, 2015). In its current projects, the technology is designed to scan a large number of geologically promising areas.
Among the multitude of scientific and applied problems solved with the help of hyperspectral sensing data gathered by drones, the main ones are the detection of areas that are promising for the search of minerals.
These problems are well studied, but remain relevant:
- The discovery of regions that are promising for the search for new oil and gas fields;
- The detection of geological objects confined to deposits of polymetallic ores, uranium ore, or diamonds in other minerals;
- The search for freshwater resources in arid areas.
In the next decades, the technology can also be practiced to transmit signals to the long-term monitoring centers at high altitudes in case of emergency. Such a perspective is especially useful for industrial minerals identification (Dar, Bukhari, and Yousuf, 2017). The need for the last is increasing due to the constant growth of the planet’s population.
Thus, hyperspectral images regarding remote sensing allow identifying the composition and the state of the minerals. They determine the geological structure and discern the chemical components of water. It signifies one of the most hopeful methods of sensing regarding economic and technological progress. The advantage of hyperspectral data is the ability to select the most informative features for each monitoring task. The method of forming such a system is a severe separate problem that is necessary for consideration in future studies. The advantages of the approach are obvious: low weight of the equipment; compatibility with most professional drones; high performance and accuracy. However, quite a reduced amount of energy with which the current drones are charged represents the limitation and requires further technological development. Quite soon drones will be able to be in the air for an unlimited amount of time due to their design, which supposes the drone functionality to be combined with the possibility of renewing batteries in the air due to renewable energy sources.
Dar, A. M., Lasitha, S., Bukhari, K., & Yousuf, M. (2017). Delineating deep basement faults in Eastern Dharwar Craton through systematic methods of geophysics and remote sensing vis-à-vis the concerns of moderate seismicity. J Geogr Nat Disast, 7(184), 2167-0587.
Gupta, R. P. (2017). Remote sensing geology. Springer.
Jordan, B. R. (2015). A bird’s-eye view of geology: The use of micro drones/UAVs in geologic fieldwork and education. GSA Today, 25(7), 50-52.
Khan, S. D., & Jacobson, S. (2008). Remote sensing and geochemistry for detecting hydrocarbon microseepages. Geological Society of America Bulletin, 120(1-2), 96-105.
Landgrebe, D. A. (2005). Signal theory methods in multispectral remote sensing (Vol. 29). John Wiley & Sons.
Sabins, F. F. (1999). Remote sensing for mineral exploration. Ore Geology Reviews, 14(3-4), 157-183.
Sabins, F. F. (2007). Remote sensing: Principles and applications. Waveland Press.
Villiger, E. J. (2008). Radar and multispectral image fusion options for improved land cover classification (Doctoral dissertation).