This study is based on extract lineament of the study area. Mosaic work for three satellite scenes for ( Landsat-8 band 8) mid Iraq, has been used to the region located between longitudes 44?? 35′ 22.27″ East to 44?? 59′ 3.97″ East and latitudes 32?? 48′ 57.49″ North to 34?? 6′ 27.27″ North. It considers as a very important structural and geological indicator to determine general and local tectonic trends and fracture zones. Throughout this study, was used for auto and manual extraction under user- suggested parameter values within the PCI Geometca software. Three geospatial analyses are applied in order to evaluate the lineaments, these are: long, density and trend. Using the manual method found that the lineament density concentrate on the N-S direction. From comparison between manual and automated methods, notes that the lineament density would be the largest manual method. Focused features extracted at the faults zones.
Keywords: Lineament extraction, mid Iraq, Landsat-8, lineament density
Lineaments are defined as mapable linear surface features, which differ distinctly from the patterns of adjacent features and presumably reflect subsurface phenomena [O’Leary, 1976]. Earth surface linear features have been study theme for geologists through many years. The old term lineament, introduced at the beginning of the 20th century. Hobbs(1911) is one of the first geologists used lineaments and realized that this features are the result of zones of weakness or structural displacement in the crust of the earth, also, Hills (1953) is one of first geologist considers lineaments as a worldwide pattern in features such as faults, fractures and major relief forms [Lattman and Nickelsen, 1958]. The importance of geologic structures especially lineaments such as joints, fractures and faults cannot be underestimated. This is because they act as reservoir both for oil, water and gas and also for the deposition of important ores. One of the main features of geological interpretation of satellite imagery has been the recognition of lineaments varying in length from a few kilometers to hundreds of kilometers [Onyedim and Ocan, 2001]. Remote sensing techniques are usually adopted in studying lineaments because they give an opportunity of synoptically studying the feature without actually coming into contact with them, especially at the regional level. Satellite remotely sensed data has been widely used as source of information for geologists to map lineaments at district and regional scales. The lineament is a mappable linear or curvilinear feature of a surface whose parts align in a straight or slightly curving relationship [Hung, 2005]. They may be an expression of faults, joints or other line weakness. The lineament may be has a geomorphological implication, i.e. major structural ridges, cliffs, terraces and aligned segments of a valley are typical geomorphological expressions of lineaments. Differences in vegetation, moisture content, and soil or rock composition account for most tonal contrast which are used to extract the linear feature [O’Leary ,1976]. Satellite images and aerial photographs are extensively used to delineate lineaments for different purposes, such as defining geological structures and tectonic fabrics [Neawsuparp and Charusiri, 2004]. Since satellite images are obtained from varying wavelength intervals of the electromagnetic spectrum, they are considered to be a better tool to discriminate the lineaments and to produce better information than conventional aerial photographs [Casas et al, 2000].
The study area is located in Mesopotamian Fore deep Basin, Middle Iraq between Longitudes 43??35′ 22.70″ E to 44?? 59′ 3.97″ E and Latitudes 32?? 48′ 57.49″ N to 34?? 6′ 27.27″ N, within area approximately equal 18659.45 km2 as shown in (Figure 1) and (Figure 2).The study area is located in the region east of the Euphrates River, where include East Baghdad oil field.
Geology of the study area:
The study area is part of the Arabian Basin in a regional extent and specifically within Mesopotamian Fore deep Basin. Locating this studied area in the Arabian Peninsula Basins map prepared by the U.S.Geological Survey [Pollastro, 1999] with special emphasis on Iraqi region of northeastern Arabian Peninsula figure (1), The study area could be grading tectonically northeastward toward the Zagross Fold Belt and westward toward the boundary with the Widian Basin of Interior Platform while its southward extensions is the Mesopotamian Fore deep Basin that contain deposits of the Tethys ocean during the Jurassic and Cretaceous Periods. Tethys Ocean was of mainly dysoxic’ anoxic palaeo environments along the equator and of tectonically unrest [Sharland, 2001] that permitted preservation of high organic matters and development of highest world oil and gas reserve in the Arabian Region. See (Figure 1).
(Figure 1): A) Location map of Iraq showing basins, oil fields and showing study area; B) Landsat-8 false color composite image covered study area; C) Geological map modified from [Geosurv-Iraq, 1994]
Materials and methodologies
Landsat-8 Operational Land Imager (OLI) images consist of nine spectral bands with a spatial resolution of 30 meters for bands 1 to 7 and 9. The resolution for band 8 (panchromatic) is 15 meters [Landsat-8 metadata file]. Landsat-8 (OLI) was chosen because of its spectral discrimination of a variety of characteristics that are required for the study. A total of three Landsat scenes wholly cover the study area. Band 8 (0.50 – 0.68 micrometers); panchromatic, is useful in the extraction of geological formations and rock features. Previous lineament extraction studies using Landsat had made use of TM, ETM and ETM+ sensors; the corresponding panchromatic band in OLI was used in this research. Three Landsat scenes, band 8, fully covered the study area acquired in different months and/or years. The Landsat -8 data used in the study are listed in (Table 1). It is geo-referenced to the UTM coordinate system; Zone 38 North and resampled using nearest neighbor resample technique.
Table (1) ETM+ data acquisition
Raw Path Date of acquisition
36 169 16 / 4 / 2000
37 169 18 / 3 / 2001
37 168 25 / 4/ 2000
The main flow chart which is applied for the lineament extraction and analysis is given in (Figure 2).
Input Data (Landsat-8 band 8 (OLI))
Sobel-Kernels directional filter
E-W N-S NE-SW NW-SE
Convert the extracted maps to a shape file format and export them to the ArcGis program
Calculate the length of the lineaments
Draw a histogram frequency for each map
Calculate density and orientation of lineaments
(Figure 2) Flowchart shows steps lineaments extraction and analysis
The first step of the methodology is selection of initial input data for lineament extraction. Although the lineaments can be extracted from several data such as aerial photographs, geophysical data etc., in this study the satellite image is preferred for the application.
The second step of the methodology is extraction of lineaments from satellite images and final map generation. This is the main step in the application. Lineament extraction in this study is performed in two ways:
‘ Manual lineament extraction.
‘ Automated lineament extraction.
Manual lineament extraction
In manual extraction method, the lineaments are extracted from satellite image by using visual interpretation. The lineaments usually appear as straight lines or ‘edges’ on the satellite images which in all cases contributed by the tonal differences within the surface material. The knowledge and the experience of the user is the key point in the identification of the lineaments particularly to connect broken segments into a longer lineament [Wang, 1990]. Some general features, however, help to identify the lineaments can be listed as follows as already described in the literature:
‘ Topographic features such as straight valleys, continuous scarps.
‘ Straight rock boundaries.
‘ Systematic offset of rivers.
‘ Sudden tonal variations.
‘ Alignment of vegetation.
According to [Koike, 1995] a continuous straight valley is the most helpful feature as a primary identification criterion in image processing for lineaments because a satellite image has no direct information on the topography of the area.
There are several image enhancement techniques that can contribute to manual lineament extraction. In this study will be used the filtering operation in the preparation of the final lineament map.
One of the characteristic features of the satellite images is a parameter called spatial frequency which is defined as the number of changes in brightness value per unit distance for any particular part of an image [Sarp, 2005]. If there are very few changes in brightness value over a given area in an image, this is referred to as a low-frequency area. Conversely, if the brightness values change dramatically over short distances, this is an area of high frequency detail [Jensen, 2005]. Therefore, filtering operations are used to emphasize or deemphasize spatial frequency in the image. This frequency can be attributed to the presence of the lineaments in the area. In other words, the filtering operation will sharpen the boundary that exists between adjacent units.
The main disadvantage of the filtering method is that it cannot effectively extract lineaments in low-contrast areas where features extended parallel to the sun directions and in mountain shadows [Koike, 1995].
In this study, Directional Gradient-Sobel filter is applied to the Landsat-8 (OLI) band 8 in N-S, E-W, NE-SW and NW-SE directions to increase frequency and contrast in the image. Directional filtering has been used to enhance, extract and classified the oriented lineaments of the study area. Directional filters are applied to image using a convolution process by mean of constructing a window normally with a (3??3) pixel box of Sobel- kernels filters (Table 2). This type of filter was used in order to get a high accuracy in extraction of oriented lineaments because the directional nature of Sobel-kernels generate an effective and faster way to evaluate lineaments in four principal directions [suzen, 1998]. Four filtered images have been produced by ENVI5.1 Imagine software related to the directions N-S, E-W, NE-SW and NW-SE.
Table (2) Sobel – kernels in four principle directions
N-S NE-SW E-W NW-SE
-1 0 1 -2 -1 0 -1 -2 -1 0 1 2
-2 0 2 -1 0 1 0 0 0 -1 0 1
-1 0 1 0 1 2 1 2 1 -2 -1 0
The results of the Sobel are given from (Figure 3) for four main directions N-S, E-W, NE-SW and NW-SE. Four maps are prepared from these images.
(Figure 3) Sobel filtered images: A) N-S direction, B) NE-SW direction, C) NW-SE direction; D) E-W direction
The result lineament map for Sobel filters and its frequency histogram is shown from (Figure 4). The number of the lineaments identified in these four maps is considerably different.
(Figure 4) Lineaments maps: A) N-S direction, B) NE-SW direction, C) NW-SE direction; D) E-W direction
As shown in (Figure 5), It has been noticed that the (N-S) lineament map have higher number and (E-W) have higher length compared with the other. According to the parameters values which are used in this study, the maximum length of the lineaments is (20 Km) recorded in the (E-W) direction. In addition, the maximum frequency of lineaments is (2143) recorded in the (N-S) direction which is about (26 %) of the final map.
(Figure 5) Frequency distribution and basic statics at four principle directions of Sobel-kernel filter
Automated Lineament Extraction
The main advantages of automated lineament extraction over the manual lineament extraction are its ability to uniform approach to different images; processing operations are performed in a short time and its ability to extract lineaments which are not recognized by the human eyes.
Available software’s provide different algorithms for automated extraction. Three common algorithms are Hough transform, Haar transform and Segment Tracing Algorithm (STA) [Ko??al, 2004].
The Hough transform is a technique which can be used to separate features of specific shape within an image. It is required that the specific feature must be defined in some parametric form. The Hough transform is most commonly used for the detection of lines, circles, ellipses, etc. The main advantages of the Hough transform are that it is relatively unaffected by gaps in lines and by noise [Wang, 1990].
Haar transform used by Majumdar and Bahattacharya (1988) for extraction of linear and anomalous patterns in the image. This method provides a domain in which a type of differential energy is concentrated in local regions. The transform has both low and high frequency components and therefore can be used for image enhancement [Ko??al, 2004].
The Segment Tracing Algorithm (STA), which is developed by [Koike, 1995], is a method to automatically detect a line of pixels as a vector element by examining local variance of the gray level in a digital image. The automated lineament extraction in this study is performed by the LINE module of Geomatica software. The logic of this method is similar to STA. A brief explanation of the algorithm of this module will be given here. This information is provided from the Geomatica users’ manual (2014).
The automated lineament extraction operations are applied on Landsat-8 (OLI) scene by using PCI Geomatica 2014 software line option. Band 8 of the image with a spatial resolution 15*15 meter is selected for automated lineament extraction considering the purpose of this study; since this band is useful for discrimination of lineaments and other geological features such as mineral and rock types and is also sensitive to vegetation moisture content [Sabins, 1996].
The extraction process is manipulated changing the six parameters. Several lineament maps are generated using different threshold values. The most suitable threshold values are selected (below) considering these lineaments as fault lines. General properties of faults are taken into consideration such as the length, curvature, segmentation, separation and so on in order to determine the threshold values. The parameters in this application are selected as follows in (Table 3).
Table (3) Suggested parameters values
Parameters Suggested value
Filter radius (RADI) 10
Gradient threshold (GTHR) 75
Length threshold (LTHR) 30
Line fitting error threshold (FTHR) 3
Angular difference threshold (ATHR) 1
Linking distance threshold (DTHR) 40
The automatically extracted lineament map and its basic statistics are illustrated in (Figure 6) and (Figure 7) respectively. The results of the manual extraction and its basic statistics also given in these figures to compare the two maps.
The Results and Discussion
Lineaments maps and geospatial analysis
In order to obtain the most appropriate lineaments related to the tectonic setting of the studied area, optimum values for LINE modular parameters are suggested (Table 2).In manual extracted, previous four filtered images (Figure 3) are used as an input data to the line modular in order to calculate and estimate the length, orientation, numbers, and density of the lineament to each one of these input data (i.e. four filtered images). (Figure 4) demonstrates the lineaments map over the four input data with different trends. In this context, lineaments are analyzed by three process of geospatial analysis in order to extract further information related to distribute and nature of these structures. Geospatial analysis process are includes: length, density and orientation analysis.
The relationship between the lineaments in each one of the four maps in number (frequency) and lengths is shown in (Figure 5). A total of (8135) geologic lineaments (for all directions) were identified digitally. Length per unit area for each line is completely calculated digitally and then represented the value of length (in kilometer) by attributes table in the data base as a new field.
Automated and Manually Extracted
-From (Figure 7) the frequency of manually extracted lineaments is greater more than 4 times of the automatically extracted ones (8135 versus 1849). The most important factor for this is that the lineaments in manually one are shorter in length so that a few of them could be combined to form one line in automatically extracted map. Although the linking distance threshold is assigned as 1200 m (40 pixels), the program could not combine segmented lines see (Figure 6).
-In addition to the frequency of the lineaments is higher in manually one, the total length of all lineaments is still higher than the lineaments (10332.83 km versus 2385.54 km) identified by automated methods. This feature is best illustrated by the mean lengths of automatic and manual lineaments which are 1.29 km and 1.27 km, respectively.
-Spatial distribution of the lineaments in both maps (Figure 6) is considerably different. In the manually one the frequency of the lines seems to be higher in the eastern and northeaster parts of the area particularly in the close vicinity of Baquba and east of Baghdad. In the manual one, on the other hand, In addition to the areas mentioned, along the Tigris and Euphrates rivers and their extensions (that fits the Tikrit-Amara Fault Zone and Ramadi-Musayib Fault Zone).
-The pattern of the two maps in general look similar.
-Length of the maximum lineament detected by automated one is 19.35km which is a proper length for the faults in the area. This length, however, is 20.01 km for manual one which is quite reasonable.
(Figure 6) Lineament maps: A) Automatically extracted; B) Manually extracted.
(Figure7) basic statistics: A) Automatically extracted; B) Manually extracted lineaments.
This analysis calculates the frequency of the lineaments per unit area [Hung et al., 2000], and then produce a map showing concentrations of the lineaments over unit area. In this study, the lineament density is created by spatial analyst tool in (ArcGis 10.2.2) program by counting lines digitally per unit area (number of lines/km2) and then plotted in the respective grid centers and contoured using the same tool. Lineaments density map of the overall lineaments (the four directions) is produced and shows in (Figure 8) by grids for manually and automated. The high density of lineaments are located in the areas within inside the main structures (i. e. anticlines). Meanwhile, it is clear that most areas adjacent to the main faults has also a high density of lineaments.
(Figure 8) Lineaments Density maps of the overall lineaments: A) Automated extracted;
B) Manual extracted
Lineaments orientations are usually analyzed by rose diagram in all researches which are dealing with these structures. In manually extraction, these diagrams shows the directional frequencies of the extracted lineaments overall the specific area in the results lineaments maps for Sobel filters. A rose diagram tool from the (RockWorks16 Software) was used to derive lineament directions in the selected part in the studied area. Orientation of the lineaments for different lineament maps are compared using the rose diagrams (Figure 9) noted to four principal directions. As well the rose diagram shows four directions (i.e. NE-SW, N-S, E-W and NW-SE) are noticed but in different ranges, however, the dominance trends in the directions are include: NE-SW, NW-SE. The diagrams are prepared using the frequencies of the lineaments and therefore are not length-weighted.
(Figure 9) Rose diagrams prepared from lineaments extracted. A) N-S direction; B) NE-SW direction; C) E-W direction; D) NW-SE direction
Two diagrams in (Figure 10) show rose diagram for manually and automatically lineaments are great similarities as being concentrated in NW-SE direction. Comparison of two maps indicates that the manually and automated extracted lineament map is more similarity in terms of their segmentation, their spatial distribution and their orientation except length of the lineaments.
(Figure 10) Rose diagrams prepared from: A) Manual extraction; B) Automated extraction
This study can be efficient way for extracting and analyzing the geological lineaments over large regions with little outcrops (covered area). Combination of auto extracted lineaments with the geospatial data (length, density and trend) can determine the fracture zones.
‘ Casas, A. M., Cortes, Angel L., Maestro, A., Soriano, M.A.,Riaguas, A., Bernal, J., 2000, A program for lineament length and density analysis, Computers and Geosciences, Vol. 26, No. 9/10, P. 1011-1022.
‘ Jensen, J. R., 2005, Introductory digital image processing : a remote sensing perspective , 3nd Ed., Prentice Hall, New Jersey, P. 17-62.
‘ Hung, L.Q., Batelaan ,O. and De Smedt, F., 2005. Lineament extraction and analysis, comparison of Landsat ETM and Aster imagery. Case study: Suoimuoi tropical karst catchment, Vietnam, Proc. of SPIE Vol. 5983, 59830T, P. 1-12.
‘ Ko??al, A., 2004, A Methodology for Detection and Evaluation of Lineaments from Satellite Imagery, Ms. thesis, Middle East Technical University, 121 p.
‘ Koike, K., Nagano, S. And Ohmi, M., 1995, Lineament Analysis of Satellite Images Using A Segment Tracing Algorithm (STA), Computers and Geosciences, Vol. 21, No. 9, P. 1091-1104.
‘ Lattman, L. H. and Nickelsen, R. P., 1958. Photogeologic fracture-trace mapping in Appalachian plateau, geological notes, P. 2239-2244.
‘ Neawsuparp, K. and P. Charusiri, 2004. Lineaments Analysis Determined from Landsat Data Implication for Tectonic Features and Mineral Occurrences in Northern Loei Area, NE Thailand, ScienceAsia. Vol. 30, P. 269-278.
‘ O’Leary, D.W. J.D. Friedman, H.A. Pohn, 1976. Lineament, linear, lineation: Some proposed new standards for old terms. Geological Society America Bulletin, Vol. 87, P. 1463-1469.
‘ Onyedim, GC, Ocan OO (2001). Correlation of SPOT imagery lineaments with geological fractures in parts of ilesha area, southwestern nigeria. J. Min. Geol.,Vol. 37, No. 1, 15-22.
‘ Pollastro RM, Karshbaum AS, Viger RG 1999 ‘Map showing Geology, Oil and Gas Fields, and Geologic Provinces of the Arabian Peninsula’. US Geological Survey, Open File Report 97- 470B, Version 2.
‘ Sabins, F. F., 1996, Remote Sensing: Principles and Interpretation, 3rd Ed., W. H. Freeman a, New York, 494 pages.
‘ Sarp, G??lcan, 2005, Lineament analysis from satellite images, north-west of Ankara, thesis submitted to the graduate school of natural and applied sciences of Middle East technical university, Ankara.
‘ Sharland PR, Archer R, Cassey DM, Davies RB, Hall SH, Heward AP, Horbery AD, 2001, “Simmons MD Arabian plate sequence stratigraphy”. Gulf PetroLink, Bahrain, p 371.
‘ S??zen, M.L. and V. Toprak, “Filtering of Satellite Images in Geological Lineament Analyses: An Application to a Fault Zone in Central Turkey”, International Journal of Remote Sensing, Vol. 19, No. 6, P. 1101-1114, 1998.
‘ Wang, Jinfei, Member, IEEE, Howarth, Philip J., 1990, Use of the Hough Transform in Automated Lineament Detection, IEEE Transactions on Geoscience and Remote Sensing, Vol. 28, No. 4, P. 561-566.