1 Geomonitoring in the post-mining environment
The sustainable management of georesources is governed by a specific life cycle that begins with the allocation of mining licences and continues with the exploration and mineral extraction stage and the operation of the site before concluding with the mine closure process and the post-mining phase (1). This manner of dealing with geological resources has a huge impact on people, nature and the environment and is an extremely complex issue that involves all kinds of individual factors. It is therefore important to ensure that a continuous programme of spatial monitoring is put in place to detect the resulting environmental changes that occur both during and after the actual operational phase. These changes include the impact of mining operations on water resources, e. g., as a result of mine dewatering activities undertaken as part of regional water management schemes. As far as soil resources are concerned mining processes can result in soil movements in the form of ground subsidence and uplift and can lead to changes in vegetation and land usage. Mining-related gas emissions can also have an impact on air resources. Environmental and geomonitoring in the post-mining era calls for an integrated, spatial monitoring plan based on various datasets that have to be collected, analysed and interpreted (2, 3). These can originate from data that has been acquired during the mining phase, such as mine plans and geological and operational maps, recordings from satellite-supported multispectral and hyperspectral sensors, photo documentation from copter and aircraft flights and data recorded by in-situ sensors. These heterogeneous data records provide the basis for an understanding of the post-mining processes.
This paper presents an overview of the latest environmental and geomonitoring methods that are available for making observations from space and from the air of post mining-relevant objects that are present on the Earth’s surface and below ground level. It also discusses the subject of data integration in a comprehensive system in relation to multi-sensor observations in different spatiotemporal resolutions.
2 Satellite-supported remote sensing
Remote sensing is a technique in which specialised sensors are used to obtain data on phenomena, processes and objects from a remote distance (4). The sensors can be mounted on satellites, aircraft or copters. Satellites on different missions orbit the Earth and their sensors regularly scan the Earth’s surface and record their changes over the course of time. The satellite data are made available free of charge on online portals such as the non-commercial Copernicus programme that was set up by the European Space Agency (ESA). Copernicus concentrates on a series of Earth-observing Sentinel satellites each with its own specific objective. Sentinel 1, e. g., monitors deformation of the Earth’s surface using a synthetic aperture radar (SAR) instrument, while Sentinel 2 identifies changes in vegetation and land use.
2.1 Optical remote sensing
Optical remote sensing covers the visible, the near-infrared and the thermal-infrared range of spectra. Everything that can be observed is based on the radiation emitted by the sun being reflected on to objects. Satellite-mounted sensors record the radiation that is reflected by objects on the Earth’s surface, such as buildings, trees and bodies of water. These data are converted into information that can be used for analysing and interpreting the phenomena that occur on the surface of the Earth. One major advantage of satellite data is their massive spatial coverage. However, cloud cover can mean that not all the images recorded by the optical sensors are suitable for use.
Remote sensing can also be used to monitor post-mining activities. This means that in the years ahead it will be possible to apply preventive measures of one kind or another (5). One aspect of the environmental and geomonitoring of post-mining processes involves the observation of changes in vegetation cover based on a calculation of vegetation indices. One of these is the Normalised Difference Vegetation Index – NDVI (6), which extracts information on vegetation cover by combining surface reflections in the red and near-infrared channel. One of the relevant features here is that healthy vegetation absorbs the red component of the sunlight almost completely – it is needed for photosynthesis – and reflects a large part of the incident solar energy in the near-infrared range. The greener the plant the greater is the increase in the reflection factor from the red spectral range to the near-infrared range. Unhealthy vegetation, on the other hand, exhibits a different reflection behaviour (Figure 1).
This characteristic of a healthy vegetation was defined mathematically using the NDVI as follows (6):
Here ρNIR corresponds to the reflection value in the near-infrared wavelength range of about 800 to 900 nm and ρRED represents the reflection value in the red range of about 600 to 700 nm. The individual wavelength ranges depend on the respective satellites and their on-board sensors. The NDVI adopts values of – 1 to 1. Water bodies show values close to – 1, with the value 0 usually indicating that no vegetation is present. The healthier the vegetation, the higher the value of the NDVI.
Many examples exist where these vegetation indices have been used to monitor changes in vegetation, including in post-mining areas in China (8), the Czech Republic (9, 10), Germany (11), Poland (12), Turkey (13, 14) and the USA (15). These case studies, which used individual vegetation indices, show just how useful and effective indices of this kind can be for monitoring the vegetation cover of post-mining processes.
Another important aspect is the spatiotemporal analysis of satellite images (Figure 2), whereby changes in vegetation and in the associated moisture levels, as occurring in the context of post-mining processes, can be observed in a specific area.
A good example of this is presented in the paper by Padmanaban et al. (11), who uses the NDVI to point out two areas with negative vegetation changes and marshland formation in and around the Kirchheller Heide nature reserve.
2.2 Radar remote sensing
Radar stands for “Radio Detection and Ranging” and is based on the principle of measurement by echolocation that involves measuring the echos of a transmitted signal. The signal carriers are electromagnetic waves in the microwave range of 300 to – 1 GHz. An active imaging radar system is mounted on to an aircraft or satellite that moves along a straight flight path at a certain height above a reference surface.
Optical satellite remote sensing usually gathers Earth observation data from the reflected sunlight, this being done in a relatively short wavelength range that varies from visible light to thermal infrared. Conventional optical remote sensing is therefore very much influenced by atmospheric effects such as cloud cover, as this limits the capacity to gather information about the Earth’s surface. Unlike optical remote sensing systems radar-based systems, which have their own radiation source, can operate irrespective of the time of day the recording is taken. Radar remote sensing also uses a relatively larger wavelength that can penetrate the atmosphere. Radar images can also be taken in cloudy conditions.
One of the most widely used methods of satellite-supported radar remote sensing involves the use of a synthetic aperture. This essentially gathers two kinds of information: backscattered intensity and phase position. The intensity of the backscattered radar radiation depends on the physical characteristics of the object surfaces, e. g., roughness and geometry, and on material properties such as moisture content and dielectric characteristics. This information can be used for detecting water bodies (16) and determining soil humidity (17). The detection of expanses of water at different times, e. g., can help to identify flooded areas, as Figure 3 shows.
The phase information is used to carry out an analysis of changes in the lengths of the propagation paths of the radar signals and this serves to identify ground movements or extract the digital height model. This is based on the principle of SAR interferometry (InSAR), which is calculated from the coherent pixels of two SAR image phase differences, the so-called interferogram. The phase differences result from the image geometry and the slightly different orbit positions of the satellite at the time of the recording, on one hand, and from the object heights of the terrain and the object surface and its deformation, on the other. Atmospheric effects also cause further signal delays. The remaining component, the measurement noise, is referred-to simply as “noise”. Calculating the deformation can be done using differential SAR interferometry, e. g., which uses at least two SAR datasets recorded at different times over the same area and calculates the deformation-related phase components from the interferogram. Figure 4 shows an example of a (non-processed) differential interferogram of a mining-induced subsidence zone.
The phase-shift map, which is represented by a recurrent colour scale, shows the movement in the sight line of the satellite.
When using techniques of this kind the decorrelations (noisy area in Figure 3) produce errors in the phase processing and cause spatial data loss. For this reason time series-based procedures like Persistent Scatter Interferometry (PSI) and Small BAseline Subset (SBAS) are being developed for better management of problems of this kind and these are now being employed increasingly for monitoring ground movements. While PSI can only establish movement time series from “high quality” backscatterers with longterm-stable and strong backscattering, which are mostly only available in urban zones due to the double-bounce effect of the buildings, SBAS uses the network of highly coherent interferograms so that results can also be obtained in rural areas. In this type of terrain, as in most coalfield and former coalfield areas, the so-called “distributed scatterers” tend to be predominant, these being used in new developments in combination with “persistent scatterers” in order to identify ground movements over wide areas. An abstract explanation of the basic principles involved can be found at Crosetto et al. (18). For geomonitoring work in post-mining terrain radar remote sensing therefore plays an indispensable role due to its special capacity for identifying ground movements occurring over wide areas.
3 Copters
Unmanned aerial vehicle platforms (UAVs, copters and drones) have now entered the data acquisition industry and this technology is becoming increasingly widespread. Techniques of this kind have a number of advantages, including
- shorter and flexible recording times;
- results obtained in high resolution.
Drones are now used increasingly for monitoring and surveying mining areas. Data collection by copters, which is dynamic and certainly more reliable than traditional methods, is an effective way of providing a continuous monitoring regime for post-mining processes.
It should also be noted that as well as offering a number of advantages this monitoring method also has its drawbacks:
- It is dependent on weather conditions (wind, rainfall, temperature).
- Its use is governed by national and EU regulations.
- It relies on GNSS signals (Global Navigation Satellite System).
Using copter aircraft allows different types of sensor to be mounted on the UAV. Passive sensors based on electromagnetic radiation can be used, e. g., in the form of digital cameras. Active sensors, which emit electromagnetic waves to a target on the Earth’s surface and then collect the signals, can also be employed on this type of platform, e. g., LiDAR (light detection and ranging). The monitoring of mining processes can be undertaken using the following types of sensing device (19):
- digital cameras;
- multispectral and hyperspectral cameras;
- thermal imaging cameras, gas detectors;
- LiDAR sensors.
This system allows orthophoto maps and thermal analyses to be produced, depending on the sensor equipment (Figure 5) (20, 21, 22).
Copters provide a much higher spatial resolution than satellite images. Multispectral cameras (red, green, blue, red edge and NIR) mounted on copters also make it possible to calculate highly-accurate vegetation indices, with the centimetre-precise resolution capacity even allowing a distinction to be made between individual plant types.
Thanks to copter flights it is now possible to identify the phenomena occurring in a certain area and to do this both in 2D and also visualised in 3D space. This means that digital surface models can be produced along with the detailed depiction of building models. This technology tool is now being used increasingly in the field of Building Information Modelling (BIM). Commercially available mapping software such as Agisoft Metashape, Pix4D and ESRI’s Drone2-Map now enable a much better processing of copter images. The results obtained here deliver a very high level of spatial resolution.
4 In-situ measurements
In-situ measurements, which are taken directly at the target object at discrete locations either at a particular time or over a particular period, play a vital role in environmental and geomonitoring. Depending on the type and purpose of the in-situ sensor, which includes the spacing between individual sensing units, various geostatistical methods can be used in order to extend the phenomena present in a certain area and blend these with other methods.
By contrast, data from remote sensing devices, while collected over a wide area, are acquired obliquely to the target zone. In the case of optical satellite remote sensing and copter-based optical recording, e. g., sensor images are captured from reflected sunlight. Various indices are used to calculate the quantities of interest from the information collected at different frequencies (see section 2.1). Rather than providing area-wide information the in-situ measurements and observations generally deliver single-spot, direct measurement results that are very important for the environmental and geomonitoring of changes, including those affecting the post-mining environment. Local on-site analysis is a key part of the validation process when it comes to using indices like NDVI to derive information on the vitality of the vegetation and the stress it is undergoing. It should also be borne in mind that only the uppermost layer of vegetation can be registered by the remote sensing process. The lower vegetation layers down to ground level are invisible to nadir image recording and require additional local input by way of site inspections in order to complete the dataset.
Terrestrial measurements obtained from geodesic and mine survey work are of particular importance for the radar interferometric identification of ground movements. InSAR methods only determine movement in the sight line of the satellite, which is not relevant for most applications that require the detection of gravity-oriented three-dimensional surface movement components. Findings determined by radar interferometric means can be validated and/or corrected on the basis of terrestrially determined height changes, e. g., using levelling, and three-dimensional movements measured by GNSS.
Measurements taken by soil moisture sensors (Figure 6) are required for many of the processes that seek to calculate soil moisture levels using radar remote sensing.
Soil sensors of this kind can continuously detect the soil moisture levels present at the sensor point at different depths. The C2M2 project being run at the TH Georg Agricola University (THGA) in Bochum/Germany, e. g., also set out to use this capability to measure and derive the seasonal changes in soil moisture levels. Because of the many different determinants affecting the propagation path of the radar waves it is only possible to a limited extent to completely separate the soil-moisture signal from other factors. The link between the SAR recordings and the appropriate model for deriving soil moisture can only be made when the former are viewed together with the measurements taken over the same period.
As well as serving to verify the results of the image processing stage (the so-called “ground truth”) and to adapt the calculation model used for the various indicators, in-situ measurements play a crucial role for many of the target objects. The observation of objects that are below ground level or that are not visible for image recording from above is only possible by taking in-situ measurements. This can include, e. g., groundwater monitoring wells and the taking of samples for soil analysis and mapping purposes. As the monitoring of exposure to pollutants is one of the most important aspects of post-mining survey work it is vital that a network of regular in-situ measurement stations is established for this purpose.
It is also worth mentioning the use of mobile GIS, which enables the recording, updating, logging and display of geodata using a smartphone-based application in the field. When working outdoors the input of data increases the informative value of the target-object description by way of direct observation, something that is not possible on a map or satellite image. Geocoded recordings by mobile GIS can therefore also be used as “ground truth” (Figure 7).
In addition to the data described, which may, e. g., be saved in a form document, it is also possible to add attachments in the form of photos. One of the drawbacks of this method is the decrease in positional accuracy of the location established by the GNSS system, e. g., when working in a woodland area.
5 Data fusion for a better understanding of the system
By using methods based on satellites, copters or in-situ measurements is it possible to provide monitoring results with different spatial and temporal resolutions for various subject areas. Appreciating the post-mining area in its entirety as a complex system, leading to an understanding of the processes involved, is a precondition for the environmental and geomonitoring of the changes taking place. This makes the integration of the different monitoring components a crucial step forwards.
For mapping and quantification work in the field of environmental and geomonitoring the use of satellite remote sensing can provide information over a large area with in most cases a low spatial resolution of a few metres. When using optical satellite remote sensing the presence of cloud cover can result in a complete information loss. Here the deployment of survey copters can significantly improve the local spatial resolution (in cm). The low flight altitude of these UAVs allows them to retain their compatibility even under cloudy, and still, weather conditions. Recordings taken by copter for local survey purposes, which are in comparable spectral ranges and bands and taken over a similar period of time, can be used to complete and improve the resolution results. It is possible to capture data using copter flights within the necessary time frame when satellite recordings cannot be used, or cannot be made, due, e. g., to excessive cloud cover.
The fusion of diverse monitoring data with different temporal resolutions is also of great benefit within the time axis. The repetition cycle for optical satellites takes several days to complete, whereas copter recordings can be undertaken spontaneously. Continuous measurements are possible by deploying in-situ sensors in an object-oriented way. Validations and additions provided by in-situ measurement systems serve to confirm the derived indicators, which are used for monitoring objects to a high quality level, and the object-related information. The positional accuracy of the different types of data and the comparability of data of diverse resolution and quality is one of the greatest challenges. It may be necessary to develop local solutions in order to ensure correct data fusion from the different scaling criteria. The specific attributes of different environmental and geomonitoring techniques are listed in Table 1.
Many issues of interest arise in connection with environmental and geomonitoring in the post-mining sphere. In this particular field of monitoring there is an added complexity to be taken into account due to the interlinked nature of the resources involved. A proper understanding and associated interpretation can only be realised when the system is recognised and accounted for in its entirety. And this can only be achieved when there is access not just to the relevant body of data but also to the knowledge and experience of experts capable of ensuring that environmental and geomonitoring in post-mining terrain is implemented in a sustainable way.
6 Conclusions
In summing up it has to be recognised that the combination and integration of the different observation methods now available remains key to the application of a successful monitoring strategy aimed at the sustainable management of georesources, especially in the post-mining environment. When dealing with a complex system it is necessary to combine the available resources with different types of data obtained from satellites, survey copters and traditional in-situ sensors. Every method has its advantages and drawbacks. However, by merging what is available it is possible to release the full potential needed to build a comprehensive understanding of how to manage georesources in a sustainable way. This generates contemporary understanding and creates transparency vis-à-vis the public. Innovative environmental and geomonitoring methods are therefore an important factor for public acceptance when it comes to the sustainable management of geological resources.
References/Quellenverzeichnis
References/Quellenverzeichnis
(1) Goerke-Mallet, P.; Melchers, C.; Müterthies, A. (2016): Innovative monitoring measures in the phase of post-mining. IMWA 2016, Freiberg/Germany, pp. 570 – 577.
(2) Rudolph, T. (2020): Digital Twin – Integrated Geomonitoring Further Developed. In: Kretschmann, J.; Goerke-Mallet, P.; Melchers, C. (2020): Done for Good 2.0, Results in Post-Mining Research, A Compilation of Research Papers by the Research Center of Post-Mining, Technische Hochschule Georg Agricola University, pp. 273 – 278.
(3) Rudolph, T.; Goerke-Mallet, P.; Melchers, C. (2020). Geomonitoring in Alt- und Nachbergbau. In: zfv 3/2020, pp. 168 – 173.
(4) Wiggenhagen, M.; Steensen, T. (2021). Guide for Photogrammetry and Remote Sensing. 6th revised and extended edition, Wichmann.
(5) Kretschmann, J.; Nguyen, N. (2020): Research Areas in Post-Mining – Experiences from German Hard Coal Mining. July-December – Journal of the Polish Mineral Engineering Society, pp. 255 – 262.
(6) Rouse, J. W.; Haas, R. H.; Schell, J. A.; Deering D. W. (1973): Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. Prog. Rep. RSC 1978-1, Remote Sensing Center, Texas A&M Univ., College Station.
(7) Simmon, R. (2000): Illustration des NDVI. In: Weier, J. und Herring, D. (2000): Measuring Vegetation (NDVI & EVI) (https://earthobservatory.nasa.gov/features/MeasuringVegetation/measuring_vegetation_2.php (Online:18.10.2021).
(8) Ma, B.; Pu, R.; Wu, L.; Zhang, S. (2017): Vegetation Index Differencing for Estimating Foliar Dust in an Ultra-Low-Grade Magnetite Area Using Landsat Imagery. IEEE Access, pp. 8825 – 8834.
(9) Halounová, L.; Petruchová, J.; Junek, P. (2005). Analysis of reclaimed areas in the Northern Bohemia brown coal open cast mining areas monitored by remote sensing data using cartography tools and GIS. WSEAS Int. Conf. on Remote Sensing, Venice/Italy, 2nd to 4th November, pp. 49 – 54.
(10) Matejicek, L.; Kopackova, V. (2010). Changes in croplands as a result of large scale mining and the associated impact on food security studied using time-series Landsat images. Remote Sensing 2, pp. 1463 – 1480.
(11) Padmanaban, R.; Bhownik, A. K.; Cabral, P. (2017). A Remote Sensing Approach to Environmental Monitoring in a Reclaimed Mine Area. In: ISPRS International Journal of Geo-Information 6 (12) 401, pp. 1 – 14.
(12) Buczyńska, A.; Blachowski J. (2021). Analysis of the vegetation condition on the area of the closed Babina mine in 1989-2019 using multispectral satellite images. IOP Conference Series Earth and Environmental Science 684 012030, pp. 1 – 13.
(13) Erener, A. (2011): Remote sensing of vegetation health for reclaimed areas of Seyitömer open cast coal mine. In: International Journal of Coal Geology 86, pp. 20 – 26.
(14) Koruyan, K.; Deliormanli, A. H.; Karaca, Z.; Momayez, M.; Lu, H.; Yalçin, E. (2012): Remote sensing in management of mining land and proximate habitat. In: Journal of the Southern African Institute of Mining and Metallurgy 112 (7), pp. 667 – 672.
(15) Raval, S.; Sarver, E.; Shamsoddini, A.; Zipper, C.; Donovan, P.; Evans, D.; Chu, H. T. (2014): Satellite remote sensing-based estimates of biomass production on reclaimed coal mines. In: Mining Engineering 66, pp. 47 – 53.
(16) Hahmann, T.; Martinis, S.; Twele, A.; Roth, A.; Buchroithner, M. (2008): Extraction of water and flood areas from SAR data. The EUSAR 2008, 7th European Conference on Synthetic Aperture Radar, 2nd to 5th June 2008, Friedrichshafen/Germany, pp. 1 – 4.
(17) Esch, S.; Korres, W.; Reichenau, T. G.; Schneider, K. (2018): Soil moisture index from ERS-SAR and its application to the analysis of spatial patterns in agricultural areas. In: Journal of Applied Remote Sensing 12(2), 022206.
(18) Crosetto, M.; Monserrat, O.; Cuevas-Gonzàles, M.; Devanthéry, N.; Crippa, B. (2016): Persistent Scatterer Interferometry: A review. In: ISPRS Journal of Photogrammentry and Remote Sensing Band 115, pp. 78 – 89.
(19) Ren, H.; Zhao, Y.; Xiao, W.; Hu, Z. (2019): A review UAV monitoring in mining areas: current status and future perspectives. In: Int. J Coal Sci Technol 6(3), pp. 320 – 333.
(20) Bernsdorf, B.; Formaniuk, A.; Rudolph. T. (2020): Possibilities of a method for copter-supported gas leak detection with thermal imaging cameras in industry and hazards prevention. In: Oil Gas – European Magazine, 46th Edition, Issue 4/2020, December, pp. 13 – 20.
(21) Bernsdorf, B.; Formaniuk, A.; Rudolph, T. (2021): Ein Ansatz zur Detektion von Gaslecks mit Hilfe UAV-gestützter Wärmebildkameras. In: vfdb – Zeitschrift für Forschung, Technik und Management im Brandschutz, 1/2021, pp. 22 – 31.
(22) Bernsdorf, B.; Rudolph, T.; Goerke-Mallet, P. (2021): Über den Einsatz von Koptern im Bergbau und in der Energiewirtschaft. In: Bergbau – Zeitschrift für Rohstoffgewinnung, Energie und -Umwelt 72, S. 15 – 23.