Advertisement

Fracture Network Mapping Using Landsat 8 OLI Data and Linkage with the Karst System: a Case Study of the Moroccan Central Middle Atlas

  • Nadia HamdaniEmail author
  • Abdennasser Baali
Original Paper
  • 24 Downloads

Abstract

The central Middle Atlas, with its Mesozoic carbonate cover has undergone several tectonic and karstic phases and is characterized by a hydrological system with a complex karstification. The spatial and structural analysis of fractures and karst networks in a complex hydrological system may allow for the accurate identification of zones that favor the infiltration and recharge of overexploited groundwater. The present study aims to characterize the fracture network in the Moroccan Central Middle Atlas using multispectral satellite images from a Landsat 8 Operational Land Imager sensor for automatic lineament extraction. Our methodology focuses on a linkage between the direction, length, and density of lineaments with the characteristics of surface karst as well as the tectonic system of the study area. The remote sensing techniques used have shown their effectiveness in lineament mapping, such as principal component analysis coupled with directional filters. The resultant fracture network is oriented NE–SW, N–S, and NW–SE with a predominance of the NE–SW direction, showing a good correlation between the distribution and orientation of the lineaments and the alignments and elongations of the karstic shapes. In addition, this paper explains the tectonic origin of surface karstic shapes and the influence of tectonic and karstification on the distribution and function of the hydrological system of the Central Middle Atlas.

Keywords

Remote sensing Lineaments Automatic extraction Landsat 8 OLI Groundwater exploration 

1 Introduction

Lineaments mapping plays a major role in several geosciences applications, especially in mining and petroleum exploration [26, 44], tectonic architecture [30], and groundwater exploration [13, 24, 36, 50]. Lineaments are natural simple or composite-pattern linear or curvilinear features discernible on the Earth’s surface [30]. They may refer to faults, fractures, dykes, or joints of stratigraphic formations, rivers, and drainage areas [17, 19, 20, 30]. Other types of lineaments include the boundaries of the different land cover units, and man-made features (roads, bridges, and edges) [3, 26, 49]. Lineaments related to fractures may provide pathways for fluid flow and contribute to the formation of many economically significant oil, geothermal, and water supply reservoirs. Fracture systems control the dispersion of chemical contaminants into and through the subsurface [30]. The groundwater movements in the karstic aquifers are controlled by lineaments, noting that groundwater moves throughout the karstic canals, developing densely at intersections and in lineaments [9, 15]. In addition, hydrogeologists are successfully using lineaments to locate high-yield wells and describe more accurately subsurface structures important for regulating groundwater recharge, migration and discharge [4, 10, 13, 24, 30].

Lineament mapping using conventional methods, such as tracking faults in the terrain and aerial photography using analogue methods, are very laborious, time-consuming, difficult to replicate, and very demanding in terms of human and material resources, facts that make mapping a large area very difficult. In addition, they are subjective and dependent on the operator’s performance, which may lead to the inaccurate characterization of the lineaments or preventing the identification of all of the existing ones [3, 30]. Geographic Information Systems (GIS) and remote sensing techniques have been used for the extraction and analysis of lineaments, providing scientists with valuable information for understanding the tectonic evolution and fracture network. Several studies have described the usefulness of remote sensing for lineament mapping and the identification of geological structures, thanks to its synoptic coverage, automatic, and easily reproducible methodology [3, 13, 17, 23, 24, 32, 41, 42, 43, 44, 49]. The availability of high-resolution multispectral data and different image processing techniques allowed for the extraction of geological lineaments with greater accuracy [3]. Several automatic extraction tools have been used in the literature, such as the LINE module of the PCI Geomatica software [17, 35, 45]. Masoud and Koike [31] have developed the LINDA tool, which uses elevation data and satellite imagery data as inputs for characterizing and interpreting morphotectonic features from lineaments.

In the Central Middle Atlas, the series of Jurassic limestones and dolomites that outcrop at the level of the atlasic Causses and Anticlinal wrinkles and their juxtaposition on impermeable permo-triassic formations and intense fracturing make this region a complex geomorphological domain. However, detailed structural analysis and a better understanding of the fracture and karst networks are always lacking in this region. Accordingly, the objective of this study is to use Landsat 8 Operational Land Imager (OLI) multispectral satellite images for the extraction, analysis and interpretation of the geological lineaments, as well as the characterization of the alignment and elongation of karstic shapes. Finally, the last step consists of the identification of the relationship between the fractures obtained from lineaments and the development of karst shapes in the study area.

2 Materials and Methods

2.1 Study Area

The Middle Atlas, with a surface area of 9500 km2, is enclosed by parallels 32° 45′ and 34° 18′ N and meridians 3° 42′ and 5° 42′ W. It is bounded to the North by the South Rifain Corridor (Saiss plain), to the South by the High Atlas and the plateaus of High Moulouya valley, to the East by the Middle Moulouya valley and to the West by the Moroccan Meseta [33]. The Middle Atlas is composed of the juxtaposition of two structural units: The Middle Atlasic Causse and the pleated Middle Atlas. These are separated by a major structural line, called the North-Middle-Atlasic fault [11]. The Middle Atlasic Causse is characterized by carbonate deposits of the Lower and Middle Lias. Its structure in inclined blocks is expressed in the topography of stepped plateaus. It is affected by the Tizi n’tretten fault, oriented N 40°. The pleated Middle Atlas is the southern part of the intracontinental chain, oriented NE–SW and elongated over more than 400 km. It is organized into acute anticlinal wrinkles and large synclinal depressions (Fig. 1).
Fig. 1

Geological map of the Central Middle Atlas (Projected Coordinate System: Lambert_Conformal_Conic, modified from [48])

The climate of the Middle Atlas is a Mediterranean mountain climate type. It is characterized by an inter-annual succession of wet and dry seasons. This is a particular climate due to the influence of the Atlantic Ocean, which affects the main rainfall and humidity. The clouds coming from the west give abundant precipitations in conjunction with the Middle Atlas. This natural barrier, which constitutes the Atlas chain, creates a dissymmetry on the climatic aspect: the Atlantic side exposed to the NW is more heavily watered, while the SE side is affected by the Saharan climate in the South. The elevation of the study area ranges between 1000 and 3500 m (Fig. 2).
Fig. 2

Morphological map of the Central Middle Atlas

2.2 Data Sources and Pre-processing

The data and approach used can be summarized as shown in Fig. 3. First, the faults and karstic shapes were digitized from the geomorphological map of the Central Middle Atlas at a scale of 1:100,000 [28]. Accordingly, we implemented a GIS database of the fracture and karstic networks, including the alignments and elongations characterizing the karsts network. The alignments and the elongations of the karst shapes have been delineated manually. The alignment is defined by the straight line formed by several karst features, while the elongation represents the straight line characterizing the lengthening of the shape of each karst. The alignment was drawn manually, in such a way that the number of objects forming the alignment was as high as possible, and the line passed as close as possible to the centroid of the objects. The alignment was conducted each time within a radius of about 20 km. Afterwards, we computed the orientation (directions) of the faults, the alignments and the elongations of the karst shapes using The Rockworks software in order to validate the extraction method of the lineaments and link the fractures system and karstic network of the study area. Secondly, two OLI data images of the study area (acquired during the summer of 2016). The first one was taken in June (Path: 201, Row: 37) and the second, in July (Path: 200, Row: 37). With a high spatial (30 m), radiometric and temporal resolution, the free availability of OLI data created a great opportunity for natural resource managers, including geologists. This sensor acquires images in nine spectral bands from shorter wavelengths of visible to short-wave infrared (SWIR). The spatial resolution of this sensor varies from 15 (panchromatic band) to 30 m (multispectral bands) (Table 1). The radiometric resolution of the OLI images is 16 bits [47]. The OLI images were radiometrically calibrated, and atmospheric effects were corrected. Accordingly, the raw data in digital numbers were converted into radiance values. Then, both additive and multiplicative atmospheric effects were corrected using FLAASH Model (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes) [12]. Afterwards, the study area was clipped from a mosaic of both of the two preprocessed images.
Fig. 3

Methodology flowchart

Table 1

Characteristics of the OLI sensor

OLI spectral band

Wavelength (μm)

Spatial resolution (m)

Band 1: Aerosols (deep blue)

0.43–0.45

30

Band 2: Blue

0.45–0.51

30

Band 3: Green

0.53–0.59

30

Band 4: Red

0.64–0.67

30

Band 5: Near Infrared

0.85–0.88

30

Band 6: SWIR 1

1.57–1.65

30

Band 7: SWIR 2

2.11–2.29

30

Band 8: Panchromatic

0.50–0.68

15

Band 9: Cirrus

1.36–1.38

30

2.3 Automatic Lineament Extraction

Automatic lineament extraction (ALE) generally involves two steps, namely, contour detection and line detection [3, 13, 17, 23]. Contours represent the limits of abrupt changes in the values of neighboring pixels, often referring to lineaments. Using the LINE module of PCI Geomatica software, the extraction procedure was based on six main parameters, detailed in Table 2. This was performed by evaluating the derivatives of the OLI images, including principal component analysis, color compositions, band ratios, and directional filters. The calibration of ALE was conducted by evaluating several combinations of parameter values, respecting the literature and the ground truth, before arriving at the optimal LINE values that gave a satisfactory result. The validation of the extracted lineaments was performed qualitatively by comparing their directions to those of the existing faults and the directions of the karst alignments and elongations. Additionally, they were validated by the interpretation of their density map, according to the faults. In lineament studies, the lineament density and orientation of existing faults in the study area are parameters that are widely used in the validation of automatically extracted lineaments by remote sensing [3, 13, 17, 32]
Table 2

The different parameters of the LINE module

Step

Parameter

Unit

Order

Description

Contours detection

Filter radius parameter (RADI)

Pixel

1

Radius of the filter that will be used for contour detection. Values between 3 and 10 are recommended to avoid introducing noise. A Canny filter, known for its efficiency, is used at this level.

Edge gradient threshold (GTHR)

Unitless

2

Gradient value between 0 and 255 to be used as threshold for edge pixel detection and those remaining as background. The result will be a binary map with contour and not contour classes. Values between 10 and 70 are acceptable.

Lines detection

Curve length threshold (LTHR)

Pixel

3

Minimum length of a curve to be taken as lineament (a value of 10 is suitable). GTHR contours will be reduced by eliminating those that do not fulfill the LTHR condition.

Line fitting threshold (FTHR)

Pixel

4

Tolerance allowed in the assembling of LTHR curves to form a polyline. A curve with a smaller number of pixels than that indicated by LTHR will not be taken into account. Values between 2 and 5 are suitable.

Angular difference threshold (ATHR)

Degrees

5

Angle not to exceed between two polylines to be linked. Values between 3 and 20 are recommended. Two polylines will be linked to form the lineaments if their two end points form an angle smaller than that specified by ATHR.

Linking distance threshold (DTHR)

Pixel

6

Finally, maximum distance between two ATHR polylines to be linked. Values between 10 and 45 are acceptable

2.4 Principal Component Analysis

Principal component analysis (PCA) is an effective technique for enhancing a multispectral image for geological studies [2, 5, 40]. It allows for the reduction of the information contained in several bands, sometimes highly correlated in a new bands called principal components (PCs), with redundant data elimination, noise isolation, and enhanced targeted information [6, 16]. Several studies have used PCA in the detection of lineaments [3, 23, 38]. In this study, we suggested the use of a color composition of the first three components, which create an excellent visual image of the interpretation and increase the contrast between the various surface objects. In addition, we tested their combination with directional filters.

2.5 Band Ratios

The band rationing method is based on the spectral reflectance variations of an object across wavelengths. It consists of dividing one band by another in order to enhance the undetectable materials in the raw data based on well-defined reflection and absorption regions [39]. The band ratios (BR) images have been used in several cases of lineaments mapping [34, 35, 45], since they allow certain minerals and lithological discontinuities to be enhanced, which can be good indicators of the presence of lineaments.

2.6 Directional Filters

The application of filters (directional, laplacian, sobel, prewitt kernels) on particular bands or RGB combinations have been explored by various authors [1, 7, 22]. Directional filters (DF) improve the perception of lineaments by causing an optical effect on the image, as it is illuminated by grazing light [27]. In our study, the DF method was selected for the semi-automatic lineament extraction, because the directional nature of Sobel kernels generates an effective and faster way to evaluate lineaments in four principal directions [52].

3 Results

3.1 Characteristics of Faults and Karstic Shapes

The Central Middle Atlas area experienced colder and wetter climatic conditions during the Quaternary than nowadays [29, 46], with snowfall favoring the dissolution of carbonates. Karstic shapes developed in dolomites as well as in limestone formations [37, 53] and particularly in the dolomites of the substratum. Surface features include minor features, such as the lapiés, the collapsed dolines that outline inaccessible subsurface networks and the small karstic summit depressions. Large depressions are represented by dolines, poljes, and ouvalas. The latter correspond to different kinds of karst, which have different geometries and dimensions. To assess the origin of different shapes of karst, their linkage with the lineament network and their distribution in the Central Middle Atlas, we analyzed and interpreted their directions. The frequency of karstic shape alignments, represented by the directional rose diagram (Fig. 4b), indicates that the shapes are predominantly aligned in the directions NE–SW and N–S. The elongations of karstic shapes are shown in Fig. 5. The rose diagram of the elongations indicates that practically all shapes are elongated in the N–S direction. The result of the digitized faults in the Central Middle Atlas is shown in Fig. 6. The main faults are aligned in the directions NE–SW, which corresponds to the direction of karstic shape alignments.
Fig. 4

a The alignments map of the karstic shapes and b the corresponding rose diagram

Fig. 5

a The elongations map of karstic shapes and b the corresponding rose diagram

Fig. 6

a Map of the main faults of the Central Middle Atlas obtained from the Martin’s geomorphological map [28] and b the corresponding rose diagram

3.2 Remote Sensing Methods for ALE

The ALE was performed by the LINE module, using PCs, color compositions, band ratios, and directional filters as inputs. The process is handled by adjusting several values of the six parameters of the LINE module for each input. Accordingly, several extraction maps were generated using different threshold values. The best inputs and most appropriate threshold values that have produced reliable results were selected (Table. 3) by considering the ground truth, the spatial distribution and the orientation of the faults and karstic shapes.
Table 3

The values of the LINE parameters suggested

Parameter

Unit

Used value

RADI

Pixel

10

GTHR

Unitless

60

LTHR

Pixel

25

FTHR

Pixel

3

ATHR

Degrees

10

DTHR

Pixel

20

3.2.1 Lineament Retrieval from PCA and Color Compositions

The combination of the first three PCs (PC1, PC2, PC3) in one color composition was used as an input of the LINE module. Accordingly, 2396 lineaments of 3358 km in total length, with an average of 1.40 km, were extracted. The rose diagram shows all the directions, with two moderate directions, N–S and NW–SE, and one dominant direction, NE–SW (Fig. 7).
Fig. 7

a Map of the lineaments identified by PCA, b the corresponding rose diagram, and c the frequency distribution of lineaments of the Central Middle Atlas

3.2.2 Spectral Band Color Compositions

Color compositions of OLI spectral bands were tested for lineament retrieval. A color composition is a combination of spectral bands based on the principle of assigning image bands to three display planes of three primary colors: red, green, and blue. In this study, we assigned the OLI spectral bands, 3, 4, and 5, to the blue, green, and red channels, respectively, giving a composition in standard false colors. This combination highlights the edges of the geological formations, vegetation, hydrographic network, and geological anomaly zones. Following the integration of this color composition in the LINE module, 4125 lineaments were extracted, with a total length of 5849 km, an average of 1.41 km and a maximum length of 8 km. The computed rose diagram clearly indicates all directions with three moderate directions, N–S, NW–SE, and E–W, and one dominant direction, NE–SW (Fig. 8).
Fig. 8

a Map of the lineaments identified from RGB-543 image, b the corresponding rose diagram, and c the frequency distribution of lineaments of the Central Middle Atlas

3.2.3 Band Ratios Analysis

After a review of the most important band ratios in the literature, the color compositions of some BRs were tested. Finally, the RGB color composition using the BRs, 6/7, 3/4, and 5/6, was chosen. The BR 6/7 is known by its discrimination of minerals rich in Al–OH, Fe–OH, and Mg–OH, which can be good indicators of the presence of water along fractures [14]. The BR 3/4 give a good discrimination of vegetation density, while the 5/6 ratio displays the disturbed areas in dark or black tones, such as shading of the abrupt changes of slope and illumination areas [23, 30, 54]. The number of lineaments obtained by this method is 4952, and their total length is 7101 km, with an average of 1.43 km and a maximum length of 7.10 km. The rose diagram has two dominant directions, NE–SW and NW–SE, as well as other directions, more pronounced compared to the previous methods (Fig. 9).
Fig. 9

a Map of lineaments identified from the RGB image: 6/7, 3/4, 4/5; b the corresponding rose diagram; and c the frequency distribution of lineaments of the Central Middle Atlas

3.2.4 Directional Filters Analysis

The application of sobel DFs 5 × 5 to the band 5 (NIR) may enhance the image discontinuities, corresponding to the lineaments. The near-infrared band allows the structural details to be highlighted [21]. DFs are applied to this band in the following directions: N–S, NE–SW, E–W, and NW–SE (Table. 4). This semi-automatic lineament extraction method leads to the retrieval of 3016 lineaments, with a total length of 4567 km, an average of 1.51 km and a maximum length of 9.48 km. The rose diagram shows that the dominant direction is NE–SW, while the NW–SE direction is slightly represented, the N–S direction is moderately represented and all other directions are poorly represented (Fig. 10).
Table 4

The Sobel directional filter 5 × 5 matrices

N00

N45

− 1.0000

− 1.0000

0.0000

1.000

1.000

− 1.4142

− 1.4142

− 0.7071

0.0000

0.0000

− 1.0000

− 1.0000

0.0000

1.000

1.000

− 1.4142

− 1.4142

− 0.7071

0.0000

0.0000

− 1.0000

− 1.0000

0.0000

1.000

1.000

− 0.7071

− 0.7071

0.0000

0.7071

0.7071

− 1.0000

− 1.0000

0.0000

1.000

1.000

0.0000

0.0000

0.7071

1.4142

1.4142

− 1.0000

− 1.0000

0.0000

1.000

1.000

0.0000

0.0000

0.7071

1.4142

1.4142

N90

N135

− 1.0000

− 1.0000

− 1.0000

− 1.0000

− 1.0000

0.0000

0.0000

− 1.0000

− 1.4142

− 1.4142

− 1.0000

− 1.0000

− 1.0000

− 1.0000

− 1.0000

0.0000

0.0000

− 1.0000

− 1.4142

− 1.4142

0.0000

0.0000

0.0000

0.0000

0.0000

0.7071

0.7071

0.0000

− 0.7071

− 0.7071

1.000

1.000

1.000

1.000

1.000

1.4142

1.4142

0.7071

0.0000

0.0000

1.000

1.000

1.000

1.000

1.000

1.4142

1.4142

0.7071

0.0000

0.0000

Fig. 10

a Map of the lineaments identified by the filter Sobel 5 × 5, b the corresponding rose diagram, and c the frequency distribution of lineaments of the Central Middle Atlas

3.2.5 Combining PCA and Directional Filters

In this section, we have combined PCA and DFs in order to assess the benefit of their synergy. The ALE was conducted using the four directional filters, previously applied during the DF analysis and the PC1. The rose diagram shows a dominance of the NE–SW direction, followed by the NW–SE direction. By contrast, other directions are poorly represented (Fig. 11).
Fig. 11

a Map of the lineaments identified by PCA and DFs combined, b the corresponding rose diagram, and c the frequency distribution of lineaments of the Central Middle Atlas

4 Discussion

Several studies have used PCA in the detection of lineaments. The comparison of five different enhancement techniques (average value of all bands, PCA, BR, histogram equalization, and high-pass filter) showed that the PCA is more efficient in the identification of lineaments [23]. Paganelli et al. [38] used the PC2 for mapping lineaments. Adiri et al. [3] showed the usefulness of the PC5 of an OLI image for mapping lineaments, thanks to its highlighting of the valleys with bright pixels. In the present study, the first three PCs, containing the maximum uncorrelated information, were used for ALE in order not to miss certain types of lineaments accompanied by a complex-specific land cover, especially hydrogeological lineaments related to karstic shapes or the recharge zones of aquifers.

Using the various processes applied to the Landsat 8 image data, notably PCA, color compositions, band ratios, and directional filters, gave satisfactory results, and all methods were able to extract the major NE–SW direction while they differ slightly in the other directions obtained as well as in the spatial distribution. To exploit the contribution of all these methods, we merged all the lineaments obtained separately by these methods. This operation was done by automatically removing the redundant lineaments, with a tolerance of less than 30 m (one pixel). Accordingly, we obtained a map, called the synthesis map of lineaments (Fig. 12). This map assembled 14,489 lineaments of variable sizes, varying from a few meters to a few kilometers in length with a total length, of 28,978 km. These obtained lineaments showed a directional rose diagram dominated by the NE–SW direction (Fig. 12).
Fig. 12

a Synthesis map of the lineaments identified by PCA, RGB, BR, and colored composition; b the corresponding rose diagram, and c the frequency distribution of lineaments of the Central Middle Atlas

Compared to the lineament map obtained from the combination of PC1 and DF, the synthesis map showed a lower number and length of lineaments, but with a wider spatial distribution, considering that the lineament map resulting from the combination of PC1 and DF gathers 21,968 lineaments, with lengths varying from a few meters to tens of kilometers and a total length of 36,780 km. On the basis of the orientation of the faults and karsts, the analysis of the results indicated that the lineaments extracted from the color compositions and band ratios as well as the synthesis map are over-distributed across all zones, and their orientations do not correlate perfectly with the faults and karst elongation and alignment. These methods tend to map lineaments that are not necessarily hydrogeological and do not correlate well with the existing faults, as shown by the rose diagrams obtained. This can be explained by the sensitivity of the methods to the land cover and land use, such as lithological units, agricultural soils, steep slopes, and vegetation cover [3, 4, 25]. This issue may be alleviated by the use of SAR radar data, which are more sensitive to geomorphology than land cover, as in the case of optical images [3]. On the other hand, the use of PCs in color composition or combined with DF give results that are less distributed in space and thus more reliable lineaments. The comparison of the obtained lineaments with the ground truth, including the spatial distribution and orientation of faults, as well as the elongation and alignment of karstic shapes, has shown that the ALE, using the combination of PC1 and DF method, outperformed the other methods. In general, the ALE methods used showed that the lineaments in the Central Middle Atlas follow two dominant directions, NE–SW and NW–SE, with the presence of other directions, including N–S.

The density of the lineaments resulting from the combination of the PC1 and DF method correlates well with the major faults of the study area, showing high densities around the faults, following the predominant direction of the faults, NE–SW. This indicates a low to medium density, mainly in the middle atlasic Causse, and medium to high density, distributed in the pleated Middle Atlas and SW of the middle atlasic Causse (Fig. 13).
Fig. 13

The map of the lineament density in the Central Middle Atlas

Concerning the karstic shapes in the Central Middle Atlas, their alignments are oriented to the NE–SW and N–S directions, while their elongation showed only the N–S direction. Their analysis showed that their presence may not be linked to a specific density of lineaments, considering that some karst shapes are located in areas with low lineament densities, but overall they show a good correlation with the lineaments (Figs. 4, 5, and 13). In addition, their elongation does not correlate with the major tectonic system (NE–SW and NW–SE). Nevertheless, the alignment of karstic shapes showed a good correlation with the NE–SW system. Overall, the presence of karst shapes on the scale of the Central Middle Atlas seems to provide information as much on the alignments as on the elongation plane.

The NE–SW direction is clearly dominant in all lineament analyses and in the frequency of the main faults in the Central Middle Atlas (Fig. 5). The N–S direction also appears in all lineament analyses but with a medium frequency. However, this direction clearly dominates the rose diagram of the elongations of the karstic shapes, which gives it a valuable role that should be linked to the Middle Atlas tectonic.

If the carbonate rocks of the Mesozoic cover, which can be karstifiable (limestone) or unkarstifiable (dolomite), had karstic surface shapes, whose characteristics (elongations and alignments) follow those of the major lineaments and major faults, the origin of the karst would be tectonic. This means that tectonics are still the initial driving forces that activate the karstic phenomena in relatively cold climates and the circulation of water through the crushed zones. The dominance of the two directions suggests that karstification has taken a place during a geological era when the N–S direction faults have played a part in the Middle Atlas tectonics, controlled by the replay of the major NE–SW faults [8, 9, 18, 51]. This era could not be other than the Quaternary, where the rapprochement of Europe and Africa could have activated these faults of the N–S direction and consequently, the karstic phenomena. These faults, qualified as distension during the major fault re-games (NE–SW), constitute high hydrological potential zones, which should be confirmed by hydrogeological data, such as hydraulic conductivity. The karstic shapes in the depth are often less described in the karstic regions. However, the tectonic and structural analysis using lineament mapping may help to better understanding the characteristics of the karst groundwater system. The lineaments are good indicators of the presence of karsts. However, the carbonates lithology should be taken into consideration. Thus, the combination of lithological mapping using satellite data and lineament mapping may be useful for characterizing the karst network. Their presence in the form of a corridor following the faults is certain. In addition, the subsurface karstic shapes, created by fracturing and/or faults, have a high potential for groundwater circulation in the karstic aquifer [9]. This structural analysis, extracting hydrogeologically significant lineaments and better understanding the tectonic evolution of the area could provide useful tools for hydro-geologists, reliable resources management and development planning.

5 Conclusions

Lineament mapping is an essential component in geological studies, particularly for understanding tectonics. Considering that the network of lineaments represents the preferred pathways for groundwater circulation, a quantitative study of the network of fractures, taking into account the direction, length and density of fracture and karstic networks, allows for better groundwater exploration and management using satellite data. The OLI sensor data have demonstrated their usefulness for fracture network mapping due to their high spatial and spectral resolution and high quantization. The geomorphology of the Central Middle Atlas is the result of several geological phenomena, such as tectonics during alpine orogenesis. Using satellite-based automatic lineament extraction and the synthesis of alignments and elongations of surface karstic shapes, the study of geological lineaments representing the surface reflections of discontinuities has shown that the main directions of lineaments conform to those of the Middle Atlas structure, NE–SW and NW–SE, in addition to the N–S direction. The presence of karstic shapes in karstifiable or unkarstifiable carbonate rocks, with the same characteristics as the lineaments and structure of the Middle Atlas, explains the important role played by the tectonics in karstification, mainly by crushing rocks. This could specify the geochronology of karstification, which is closely related to the appearance of distension faults (N–S). These sub meridian (N–S) faults appeared during the sinister replay of the main faults (NE–SW) of the Middle Atlas, following the rapprochement of the African-European continents during the Quaternary. In such a complex hydrogeological context, these karstic shapes, with faults and submeridian (N–S) fractures may constitute high hydrological potential zones.

Notes

Acknowledgements

The authors would like to acknowledge the Faculty of Sciences Dhar Mahraz for its financial and logistical support. The authors also would like to thank the U.S. Geological Survey (USGS) for providing; free of charge the Landsat 8 OLI data. Finally, would like to thank the anonymous referees for their consistent reviewing and remarks.

References

  1. 1.
    Abdullah A, Nassr S, Ghaleeb A (2013) Remote sensing and geographic information system for fault segments mapping a study from Taiz area, Yemen. J Geol Res 2013:1–16.  https://doi.org/10.1155/2013/201757 CrossRefGoogle Scholar
  2. 2.
    Adiri Z, El Harti A, Jellouli A, Maacha L, Bachaoui E (2016) Lithological mapping using Landsat 8 OLI and Terra ASTER multispectral data in the Bas Drâainlier, Moroccan Anti Atlas. J Appl Remote Sens 10(1):016005 14.  https://doi.org/10.1117/1.JRS.10.016005 CrossRefGoogle Scholar
  3. 3.
    Adiri Z, El Harti A, Jellouli A, Lhissou R, Maacha L, Azmi M, Zouhair M, Bachaoui E (2017) Comparison of Landsat-8, ASTER and Sentinel 1 satellite remote sensing data in automatic lineament extraction: a case study of Sidi Flah-Bouskour inlier, Moroccan Anti Atlas. Adv Space Res 60(11):2355–2367.  https://doi.org/10.1016/j.asr.2017.09.006 CrossRefGoogle Scholar
  4. 4.
    Alonso C, Carla A (2011) Lineament mapping for groundwater exploration using remotely sensed imagery in a karst terrain: Rio Tanama and Rio de Arecibo basins in the northern karst of Puerto Rico. Master’s thesis, Michigan Technological University.. https://digitalcommons.mtu.edu/etds/309. Accessed 2 Jan 2019
  5. 5.
    Amer R, Kusky T, Ghulam A (2010) Lithological mapping in the Central Eastern Desert of Egypt using ASTER data. J Afr Earth Sci 56:75–82.  https://doi.org/10.1016/j.jafrearsci.2009.06.004 CrossRefGoogle Scholar
  6. 6.
    Amer R, Kusky T, El Mezayen A (2012) Remote sensing detection of gold related alteration zones in um Rus area, Central Eastern Desert of Egypt. Adv Space Res 49:121–134.  https://doi.org/10.1016/j.asr.2011.09.024 CrossRefGoogle Scholar
  7. 7.
    Argialas D, Mavrantza O, Stefouli M (2003) Automatic mapping of tectonic lineaments (faults) using methods and techniques of photointerpretation /digital remote sensing and expert systems (geology no. THALES Project 1174). https://pdfs.semanticscholar.org/defb/d1569f26d39d20a7ea8aaab0bd0392bc69b9.pdf. Accessed 2 Jan 2019
  8. 8.
    Aydin A (1984) Diverse Pliocène-Quaternary tectonics in a transform environment, San Francisco Bay region, California. Geol Soc Am Bull Page BM 95:1303–1317CrossRefGoogle Scholar
  9. 9.
    Baali A (1998) Genèse et évolution au Plio-Quaternaire de deux bassins intramontagneux en domaine carbonaté méditerranéen. Les bassins versants des Dayets Afourgagh et Agoulmam (Moyen Atlas, Maroc) (Doctoral dissertation, Thèse, Université Sidi Mohamed Ben Abdellah de Fès, 326p)Google Scholar
  10. 10.
    Bruning JN, Gierke JS, Maclean AL (2011) An approach to lineament analysis for groundwater exploration in Nicaragua. Photogramm Eng Remote Sens 77:509–519.  https://doi.org/10.14358/PERS.77.5.509 CrossRefGoogle Scholar
  11. 11.
    Colo G (1961) Contribution a l'étude du Jurassique du Moyen Atlas septentrional: Atlas de planches hors texte. Éd. de la Division de la géologie, Direction, Ministère, Royaume du MarocGoogle Scholar
  12. 12.
    Cooley T, Anderson G P, Felde G W, Hoke M L, Ratkowski A J, Chetwynd J H et al. (2002) FLAASH, a MODTRAN4-based atmospheric correction algorithm, its application and validation. In Geoscience and Remote Sensing Symposium, 2002. IGARSS'02. 2002 IEEE International (Vol. 3, pp. 1414–1418). IEEEGoogle Scholar
  13. 13.
    Corgne S, Magagi R, Yergeau M, Sylla D (2010) An integrated approach to hydro-geological lineament mapping of a semi-arid region of West Africa using Radarsat-1 and GIS. Remote Sens Environ 114(9):1863–1875.  https://doi.org/10.1016/j.rse.2010.03.004 CrossRefGoogle Scholar
  14. 14.
    Crippen RE (1988) The dangers of underestimating the importance of data adjustments in band rationing. Remote Sens 9(4):767–776.  https://doi.org/10.1080/01431168808954891 CrossRefGoogle Scholar
  15. 15.
    Davraz A, Karaguzel R, Soyaslan I et al (2009) Hydrogeology of karst aquifer systems in SW Turkey and an assessment of water quality and contamination problems. Environ Geol 58:973.  https://doi.org/10.1007/s00254-008-1577-5 CrossRefGoogle Scholar
  16. 16.
    Gabr S, Ghulam A, Kusky T (2010) Detecting areas of high-potential gold mineralization using ASTER data. Ore Geol Rev 38(1–2):59–69.  https://doi.org/10.1016/j.oregeorev.2010.05.007 CrossRefGoogle Scholar
  17. 17.
    Hashim M, Ahmad S, Johari MAM, Pour AB (2013) Automatic lineament extraction in a heavily vegetated region using Landsat Enhanced Thematic Mapper (ETM+) imagery. Adv Space Res 51(5):874–890.  https://doi.org/10.1016/j.asr.2012.10.004 CrossRefGoogle Scholar
  18. 18.
    Hatzfeld D, Frogneux M (1981) Intermediate depth seismicity in the Western Mediterranean unrelated to subduction of oceanic lithosphere. Nature 292:443–445.  https://doi.org/10.1038/292443a0 CrossRefGoogle Scholar
  19. 19.
    Hobbs WH (1904) Lineaments of the Atlantic border region. Geol Soc Am Bull 15:483–506.  https://doi.org/10.1130/GSAB-15-483 CrossRefGoogle Scholar
  20. 20.
    Hung L. Q, Batelaan O, De Smedt F (2005 Lineament extraction and analysis, comparison of LANDSAT ETM and ASTER imagery. Case study: Suoimuoi tropical karst catchment, Vietnam. In Remote Sensing for Environmental Monitoring, GIS Applications, and Geology V (Vol. 5983, p. 59830T). International Society for Optics and Photonics. doi  https://doi.org/10.1117/12.627699
  21. 21.
    Jourda JPR, Saley MB, Djagoua EV, Kouamé KJ, Biémi J, Razack M (2006) Utilisation des données ETM+ de Landsat et d’un SIG pour l’évaluation du potentiel en eau souterraine dans le milieu fissuré précambrien de la région de Korhogo (Nord de la Côte d’Ivoire): approche par analyse multicritère et test de validation. Télédétection 5(4):339–357Google Scholar
  22. 22.
    Kavak KS (2005) Determination of palaeotectonic and neotectonic features around the Menderes Massif and the Gediz Graben (West. Turkey) using Landsat TM image. Int J Remote Sens 26(1):59–78.  https://doi.org/10.1080/01431160410001709994 CrossRefGoogle Scholar
  23. 23.
    Li N (2010) Textural and rule-based lithological classification of remote sensing data, and geological mapping in Southwestern Prieska Subbasin, Transvaal Supergroup, South Africa. In: Thesis presented in at the Faculty of Earth Sciences the Ludwig Maximilian University Munich, GermanyGoogle Scholar
  24. 24.
    Maina MM, Tudunwada Y (2017) Lineament mapping for groundwater exploration in Kano state, Nigeria. Int J of Adv in Agri & Envir Engg 4(1):226–229. http://ijaaee.iicbe.org/index.php?id=18. Accessed 2 Jan 2019
  25. 25.
    Mansour M, Ait Brahim L (2005) Utilisation de la télédétection pour l’analyse de la fracturation du domaine interne rifain (Maroc): relation avec la répartition des sources. Télédétection 5(1–2-3):95–103Google Scholar
  26. 26.
    Marghany M, Hashim M (2010) Lineament mapping using multispectral remote sensing satellite data. Res J Appl Sci 5:126–130.  https://doi.org/10.3923/rjasci.2010.126.130 CrossRefGoogle Scholar
  27. 27.
    Marion A (1987) Introduction to Image Processing. Springer US, Springer Science+Business Media, Dordrecht, p 314.  https://doi.org/10.1007/978-1-4899-3186-3
  28. 28.
    Martin J, (1973) Carte géomorphologique du Moyen Atlas central au 1/100 000, 5 feuilles El Hajeb-Sefrou Ain Leuh-Boulmane-Kerrouchène.Notes et mémoires du service géologique. Maroc, no. 258Google Scholar
  29. 29.
    Martin J (1981) Le Moyen Atlas central, étude géomorphologique. Published by Service Géologique du Maroc, Rabat Morocco. Notes et Mémoire n° 258 et 258 bis, p 447Google Scholar
  30. 30.
    Masoud AA, Koike K (2006) Tectonic architecture through Landsat-7 ETM+/SRTM DEM-derived lineaments and relationship to the hydrogeologic setting in Siwa region, NW Egypt. J Afr Earth Sci 45(4–5):467–477.  https://doi.org/10.1016/j.jafrearsci.2006.04.005 CrossRefGoogle Scholar
  31. 31.
    Masoud AA, Koike K (2017) Applicability of computer-aided comprehensive tool (LINDA) and shaded digital elevation model for characterizing and interpreting morphotectonic features from lineaments. Comput Geosci.  https://doi.org/10.1016/j.cageo.2017.06.006
  32. 32.
    Meshkani SA, Mehrabi B, Yaghubpur, Sadeghi M (2013) Recognition of the regional lineaments of Iran: using geospatial data and their implications for exploration of metallic ore deposits. Ore Geol Rev 55:48–63.  https://doi.org/10.1016/j.oregeorev.2013.04.007 CrossRefGoogle Scholar
  33. 33.
    Michard A (1976) Eléments de géologie marocaine. Published by Service Géologique du Maroc, Rabat, Morocco. Notes et Mémoire No. 252, p 408Google Scholar
  34. 34.
    Mwaniki MW, Matthias MS, Schellmann G (2015a) A comparison of Landsat 8 (OLI) and Landsat 7 (ETM+) in mapping geology and visualizing lineaments: a case study of central region Kenya. Int Arch Photogramm Remote Sens Spat Inf Sci XL-7/W3:897–903.  https://doi.org/10.5194/isprsarchives-XL-7-W3-897-2015 CrossRefGoogle Scholar
  35. 35.
    Mwaniki MW, Matthias MS, Schellmann G (2015b) Application of remote sensing technologies to map the structural geology of central region of Kenya. IEEE J Sel Topics Appl Earth Observ Remote Sens 8(4):1855–1867.  https://doi.org/10.1109/JSTARS.2015.2395094 CrossRefGoogle Scholar
  36. 36.
    Nanda S, Annadurai R, Barik KK (2017) Geospatial decipherment of groundwater potential of Kattankolathur block of Tamil Nadu using MCDM techniques. Remote Sens App Soc Env 8:240–250.  https://doi.org/10.1016/j.rsase.2017.10.002 CrossRefGoogle Scholar
  37. 37.
    Nicod J (1972) Pays et paysages du calcaire. In: Norois, no. 79, Juillet-Septembre 1973. pp. 566–568. www.persee.fr/doc/noroi_0029-182x_1973_num_79_1_3316_t1_0566_0000_1
  38. 38.
    Paganelli F, Grunsky E, Richards J, Pryde R (2003) Use of radarsat-1 principal component imagery for structural mapping: a case study in the buffalo head hills area, northern 11 Central Alberta, Canada. Can J Remote Sens 29(1):111–140.  https://doi.org/10.5589/m02-084 CrossRefGoogle Scholar
  39. 39.
    Pour AB, Hashim M (2011) Identification of hydrothermal alteration minerals for exploring of porphyry copper deposit using ASTER data, SE Iran. J Asian Earth Sci 42:1309–1323.  https://doi.org/10.1016/j.jseaes.2011.07.017 CrossRefGoogle Scholar
  40. 40.
    Pour AB, Hashim M (2012) The application of ASTER remote sensing data to porphyry copper and epithermal gold deposits. Ore Geol Rev 44:1–9.  https://doi.org/10.1016/j.oregeorev.2011.09.009 CrossRefGoogle Scholar
  41. 41.
    Pour BA, Hashim M (2014) Structural geology mapping using PALSAR data in the Bau gold mining district, Sarawak, Malaysia. Adv Sp Res 54(4):644–654.  https://doi.org/10.1016/j.asr.2014.02.012 CrossRefGoogle Scholar
  42. 42.
    Pour BA, Hashim M (2015a) Structural mapping using PALSAR data in the central Gold belt, Peninsular Malaysia. Ore Geol Rev 64:13–22.  https://doi.org/10.1016/j.oregeorev.2014.06.011 CrossRefGoogle Scholar
  43. 43.
    Pour BA, Hashim M (2015b) Integrating PALSAR and ASTER data for mineral deposits exploration in tropical environments: a case study from Central Belt, peninsular Malaysia. Int J Image Data Fusion 6(2):170–188.  https://doi.org/10.1080/19479832.2014.985619 CrossRefGoogle Scholar
  44. 44.
    Pour BA, Hashim M, Makoundi C, Zaw K (2016) Structural mapping of the Bentong-Raub suture zone using PALSAR remote sensing data, peninsular Malaysia: implications for sediment-hosted/ orogenic gold mineral systems exploration. Resour Geol 66(4):368–385.  https://doi.org/10.1111/rge.12105 CrossRefGoogle Scholar
  45. 45.
    Qari MHT, Madani AA, Matsah MIM, Hamimi Z (2008) Utilization of Aster and Landsat data in geologic mapping of basement rocks of Arafat area. Saudi Arabia Arab J for Sci and Engg 33(1C):99–117Google Scholar
  46. 46.
    Raynal R (1961) Plaines et piemonts du bassin de la Moulouya (Maroc oriental): étude géomorphologique. Imframar, Rabat, p 617Google Scholar
  47. 47.
    Roy DP, Wulder MA, Loveland TR, Woodcock CE, Allen RG, Anderson MC et al (2014) Landsat-8: science and product vision for terrestrial global change research. Remote Sens Environ 145:154–172.  https://doi.org/10.1016/j.rse.2014.02.001 CrossRefGoogle Scholar
  48. 48.
    Saadi M (1982) Carte structurale du Maroc, 1:2,000,000. Published by Service Géologique du Maroc, Rabat, Morocco. Notes et Mémoires No. 278Google Scholar
  49. 49.
    Saadi NM, Zaher MA, El-Baz F, Watanabe K (2011) Integrated remote sensing data utilization for investigating structural and tectonic history of the Ghadames Basin, Libya. Int J Appl Earth Obs Geoinf 13(5):778–791.  https://doi.org/10.1016/j.jag.2011.05.016 CrossRefGoogle Scholar
  50. 50.
    Sahoo S, Das P, Kar A, Dhar A (2018) A forensic look into the lineament, vegetation, groundwater linkage: study of Ranchi District, Jharkhand (India). Remote Sens App Soc Env 10:138–152.  https://doi.org/10.1016/j.rsase.2018.04.001
  51. 51.
    Segall P, Pollard DD (1980) Mechanics of discontinuous faults. J Geophys Res 85:4337–4350CrossRefGoogle Scholar
  52. 52.
    Süzen ML, Toprak V (1998) Filtering of satellite images in geological lineament analyses: an application to a fault zone in Central Turkey. Int J Remote Sens 19(19):1101–1114.  https://doi.org/10.1080/014311698215621 CrossRefGoogle Scholar
  53. 53.
    Thrailkill J (1977) Relative solubilities of limestone and dolomite. Karst hydrology. AIH Memoirs 12:491–500Google Scholar
  54. 54.
    Won-In K, Charusiri P (2003) Enhancement of thematic mapper satellite images for geological mapping of the Cho Dien area, northern Vietnam. Int J Appl Earth Obs Geoinf 4(3):183–193.  https://doi.org/10.1016/S0303-2434(02)00034-X CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Laboratory of Geosystems, Environment and Sustainable Development, Faculty of Sciences Dhar MahrazSidi Mohamed Ben Abdellah UniversityFez-AtlasMorocco

Personalised recommendations