A Review of Geological Applications of High-spatial-resolution Remote Sensing Data

Geologists employ high-spatial-resolution (HR) remote sensing (RS) data for many diverse applications as they effectively reflect detailed geological information, enabling high-quality and efficient geological surveys. Applications of HR RS data to geological and related fields have grown recently. By analyzing these applications, we can better understand the results of previous studies and more effectively use the latest data and methods to efficiently extract key geological information. HR optical remote sensing data are widely used in geological hazard assessment, seismic monitoring, mineral exploitation, glacier monitoring, and mineral information extraction due to high accuracy and clear object features. Compared with optical satellite images, synthetic-aperture radar (SAR) images are stereoscopic and exhibit clear relief, strong performance, and good detection of terrain, landforms, and other information. SAR images have been applied to seismic mechanism research, volcanic monitoring, topographic deformation, and fault analysis. Furthermore, a multi-standard maturity analysis of the geological applications of HR images using literature from the Science Citation Index reveals that optical remote sensing data are superior to radar data for mining, geological disaster, lithologic, and volcanic applications, but inferior for earthquake, glacial, and fault applications. Therefore, geological remote sensing research needs to be truly multidisciplinary or interdisciplinary, ensuring more detailed and efficient surveys through cross-linking with other disciplines. Moreover, the recent application of deep learning technology to remote sensing data extraction has improved automatic processing and data analysis capabilities.


Introduction
Geology has an increasingly important role in protecting the eco-environment and providing land and mineral resources.The various categories of geological survey include: using different methods and different content for geological mapping at various scales, regional geophysical survey, regional geochemical investigation, remote sensing (RS) geological survey and evaluation, strategic mineral evaluation, regional environmental geological survey, and marine geological survey.With the development of RS technology, geologists have used these data for mineral identification and mapping, regional mapping, mineral resource exploration, mining environment monitoring, and oil and gas leakage monitoring.RS is a proven technique that can be confidently and efficiently applied in almost all disciplines of Earth science [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19], for example: (i) Geological information and remote sensing data provide information about Earth's surface structure [20].Because of the range of geological hazards and poor natural conditions, traditional geological surveys of geological hazards are very complex.Therefore, highresolution (HR) remote sensing plays an important role in geological hazard investigations.
(ii) The global use of RS for the mapping of regional structures has a long history [21,22,23].Recent advances in GIS spatial tools and the availability of various remotely sensed data have enabled the reliable delineation of topographical boundaries [24].Additionally, highspatial-resolution topographic data are a crucial prerequisite for evaluating and modeling volcanic hazards [25].
(iii) As a cutting-edge technology currently supporting the geosciences (among other disciplines), RS could play a significant role in mineral exploration research, effectively decreasing initial investments and saving time by targeting the most likely locations of ore deposits [26].
Compared to field investigations, surveying alteration minerals using RS imagery is more economical and convenient due to reduced labor costs [27].HR can provide more sophisticated geological surveys that incorporate structural interpretations, lithology, metallogenic or orecontrolling elements, regional mineral resources, and lithology-structure remote sensing interpretation maps.
Currently, there are many published reviews of RS applications in geology.Most of these reviews are related to the geological applications of hyperspectral remote sensing [28,29,30,31,32,33,34,35,36], with fewer studies involve optical and radar geological remote sensing at high spatial resolution [37,38], despite the fact that HR RS data can be beneficial for many different types of geological research.Thus, the objectives of this review are as follows: (i) analyze various HR remote sensing techniques and their applications; specifically optical satellite remote sensing and synthetic aperture radar (SAR); (ii) provide a geological application maturity analysis of HR images for reference to different research fields; and (iii) propose future research directions including multidisciplinary approaches and deep learning techniques.

Optical satellite remote sensing products and geological applications
In recent years, HR optical remote sensing satellite technology has developed rapidly, exhibiting continuously increasing resolution and more diverse imaging modes.HR satellite imagery provides a useful source for mapping the Earth's surface [39,40,41]; civil optical remote sensing satellites with a spatial resolution of better than 10 m have appeared in recent years (Table 1).The SPOT family of optical satellites better reflects the attribute information of vegetation, rock, soil, water bodies, ice bodies, and linear structures, among others, which greatly improves the accuracy of judgment and interpretation.SPOT images have been used as base maps for tracing and constructing detailed geological maps of prospect and target areas [42].Among these satellites, SPOT-5/6/7 data has played an important role in geology.Some key geology applications are as follows:

SPOT satellites
(i) Lithology: geologists have used SPOT images to extract lithologic information and predict bedrock geology mapping [43,44].
(ii) Glaciology: studying the effects of global warming on mountainous and polar glaciers is an effective use of stereo imaging capability [45,46,47].
(iv) Geological disasters: high-resolution multispectral and panchromatic images significantly help the artificial and automatic mapping of geological hazards and their comparisons for the prediction and prevention of geological hazards [51,52,53,54,55].
(v) Mine ecology: high-spatial-resolution SPOT data is an important reference for monitoring the accumulation of mine solid waste and ecological restoration in mining exploitation areas [56,57,58].

WorldView satellites
WorldView satellites have modern geo-positioning accuracy and excellent response capability, can quickly aim at the target to be photographed, and effectively performs stereoscopic imaging in the same orbit.Due to its advantages of fine spectral and spatial features, WorldView data has been widely used in: (i) Mineralogy: WorldView data is widely used to extract mineral alteration information, evaluate lithologic mapping data, and map minerals because of its high spectral resolution [59,60,61,62].
(ii) Gas detection; i.e. detection of alteration caused by gas leakage using ASTER and WorldView-2 data [63].
(iii) Topography: using the stereo imaging capability of WorldView, researchers can update topographic maps and generate DEMs [64].
(iv) Mining: Worldview-2 satellite imagery has been used to search for ancient gold mines in the Filippoi area of Macedonia, Greece [65].
(v) Glaciology: researchers have determined the 3D variation and seasonal velocity evolution of glaciers using the stereo imaging capability of WorldView data [66,67],.
Compared with TM and ASTER data, WorldView-3 data have a higher spatial, spectral, and radiation resolution and more potential for remote sensing applications related to the extraction of alteration information [68].After determining the mapping relationship between WorldView remote sensing data and the field, it is possible to clearly identify the lithofacies and belts of various strata, lithologic segments, subdivisions of intrusions, and volcanic rocks, and accurately judge the scale, morphology, and active period of various geological structures [60].Therefore, we can use WorldView data to implement HR remote sensing interpretation indicators for lithology, tectonics, and ore controlling elements.

Chinese survey satellites
The ZiYuan-3 remote sensing satellite was the first civil high-resolution stereo mapping satellite in China, which can establish digital surface models (DSMs) and DEMs with high precision.It is suitable for obtaining detailed information of micro-geomorphological features and can accurately reflect the linear image features of minor secondary fault structures [69].ZY-3 data have been widely applied in surveying and mapping, agriculture, forestry, environmental protection, disaster reduction, urban planning, and other fields [70].ZY-3 has also been used to conduct land cover mapping in surface-mined areas [40,41,71].For some high mountainous areas, ZY-3 data cannot meet the corresponding accuracy requirements, and only existing topography and geological maps can be repaired and updated.Table 2 presents some recent examples of ZY-3 satellite geological applications.A comparison of machine learning algorithms for mapping of complex surface-mined and agricultural landscapes using Ziyuan-3 stereo satellite imagery

Sentinel-2
The Sentinel-2 satellite has been widely used for agriculture and forestry planting, determining land cover and forests, monitoring glaciers, and mapping natural disasters.Moreover, the characteristics of the Sentinel-2 band have been employed to determine useful band ratios for extracting important reference minerals.Comparing the band ratio products of ferric iron, ferrous iron, laterite, gossan, ferrous silicate, and ferric oxides to a local geologic map of an imaged hydrothermal area (the Rodalquilar mining area, Cabo de Gata volcanic field, SE Spain) revealed that band ratio products supported the existing conceptual geologic model of the epithermal deposit [72].The band ratio method of Sentinel-2 data also has great potential for mapping of glacial outlines and performing velocity measurements [73,74].Additionally, the results of Hyperion/OLI and EnMAP/Sentinel-2 have been compared for mine waste monitoring [75].Because Sentinel-2 bands (bandwidth included) cover 59% of the identified useful hyperspectral bands and multispectral aims, they can play a key role in the enhancement of hyperspectral data and increase potential applications because of their complementary spatial, temporal, and spectral resolutions [76].

Synthetic aperture radar (SAR) products and geological applications
While spatial and spectral information is extracted by optical remote sensing, radar remote sensing in the microwave frequency band also provides an irreplaceable source of information for HR Earth observations with both day and night imaging capability [77] (Park, 2015).SAR images can provide a wealth of geological and mineral information such as geological structure, lithology, and hidden geological bodies, especially related to volcanic deposits, meteorite impacts, and large faults [78,79,80,81].The real time quantitative measurement of crustal deformation can be obtained by using ground Sentinel-1 images obtained at different times.In fact, the analysis of Sentinel-1 data has become routine by participating national and international agencies in seismic research, earthquake disaster assessment, and civil defense activities.The main aspects of earthquake analysis using Sentinel-1 data include: (i) Location and evaluation of hypocenters [82].
(iii) Estimation of the short-term spatial evolution of seismic sequences [84].
The use of SAR to monitor surface deformation and geological disasters has become a key issue in recent years [85].This method, which is suitable for large-scale ground deformation, has been widely used in urban surface monitoring, mine surface monitoring, dam peripheral surface monitoring, and other geological conditions.Recent studies employing SAR data with a spatial resolution of greater than 10 m are listed in Table 4.

Maturity analysis of geological applications with high resolution remote sensing data
High-spatial-resolution optical remote sensing satellite data exhibit higher resolution and more visual image information than spaceborne synthetic aperture radar data, which simplifies the interpretation and identification of geological information.Compared with optical satellite images, SAR images are stereoscopic and exhibit clear relief, strong performance, and good detection effects for terrain, landforms, and other information.The application of SAR images is highly sensitive to differences in surface morphology and roughness; thus, most radar data are still used in deserts and other low-vegetation coverage areas.The fusion of radar and optical images would significantly simplify the process of target detection and recognition and improve the accuracy of recognition by combining the advantages of both methods.
Table 5 summarizes the results of a maturity analysis of geological applications for HR remote sensing data, scaled between "low" and "high".This maturity assessment is based on a multi-standard analysis of 15 years of relevant literature from the Science Citation Index.The criteria include the quality of the high-resolution data set; i.e., the number of samples selected, the research method, the accuracy of the research results, the number of research reports, and the applicability of the field investigation.Each standard is assigned and weighted equally to obtain a final rating.a Maturity (1-5)= weighted average of quality scores for the high-resolution data set, based on the number of samples selected, the research method, the accuracy of the research results, the number of research reports, and the applicability of the field investigation.Low = 1, low-medium= 2, medium = 3, mediumhigh = 4, and high = 5.

Multidisciplinary approaches
Remote sensing as a tool for geological survey also has some limitations because remote sensing images mainly reflect surface information or shallow information, while most geological bodies, geological structures, and geological phenomena reflect three-dimensional information.The integration of geological, remote sensing and geophysical data limits the quantitative details of unknown regions, which can reduce the ambiguity of geological interpretation and improve the accuracy [86,87,88,89,90,91,92,93,94].In addition, monitoring and understanding of geological processes also requires multidisciplinary analysis and integration of remote sensing data with field observations and underground geophysical data [30].Using GIS, grid and vector data extracted from the field can be combined with remote sensing, geophysics, and geochemistry data analysis, which can comprehensively analyze different data types and their relationships, thus laying a foundation for geological analysis.This is an important trend in remote sensing geology; however, it is not always possible to benefit from these advantages and new technical insights.The combination of high-resolution remote sensing data and traditional medium-lowresolution remote sensing data (ETM and ASTER) can be used to understand geological structures and mineralization from a new perspective.It not only takes advantage of macroscopic and efficient data, but also reveals information on microstructures, strata, magmatic rock, and mineralization [95].In this way, many types of ore-forming and orecontrolling elements can be reasonably determined.Research into the spectral combination, spectral testing, and spectral inversion of remote sensing data can be used to enhance and extract specific rock types, mineralization alteration types, and mineral types, and to improve high-resolution remote sensing prospecting technology and obtain research results that can be used for rapid and low investment [96].Furthermore, combining geology with health, linking pollution research with environmental geology, or relating remote sensing to biogeochemistry are all worthy research fields [97], which could lead to better coupling of remote sensing surface information and geophysical data.Therefore, geological remote sensing research needs to be truly multidisciplinary and interdisciplinary, and geological surveys can be made more detailed and efficient through cross-linking with other disciplines.

Deep learning techniques in high-resolution remote sensing
Deep learning is a hot topic in machine learning and artificial intelligence, which imitates the mechanism of the human brain to interpret data.By training a large amount of data, deep learning can automatically obtain the relevant information and make a corresponding analysis and judgment, which can greatly improve the processing ability of high-resolution remote sensing images, thus enhancing their application value.Recently, research has been conducted on deep learning in geological hazard detection and remote sensing image recognition [98,99].Table 6 presents recently published literature on deep learning applications employing highspatial-resolution remote sensing data.It can be concluded that deep learning has good applications for image classification and scene classification but is less relevant to geology and other related fields.This is because most scene and image classification is based on artificial ground objects with geometric shape comparison rules that are easy to recognize and classify.Geological bodies are multi-type, multi-scale, multi-genesis, and multi-environment; therefore, it is very challenging to classify and recognize geological bodies from remote sensing images.

Table 1 .
Civil optical satellite data with a spatial resolution of more than 10 m used for geological remote sensing surveys in recent years.

Table 2 .
Published literature on the geological applications of ZY-3 data in recent years.

Table 5 .
Maturity a of geological applications with high-resolution remote sensing data.

Table 6 .
Recent published literature on deep learning geology applications using HR data.