Landsat 8: Utilizing sensitive response bands concept for image processing and mapping of basalts

The Egyptian Journal of Remote Sensing and Space Sciences xxx (xxxx) xxx

Contents lists available at ScienceDirect

The Egyptian Journal of Remote Sensing and Space Sciences journal homepage: www.sciencedirect.com

Landsat 8: Utilizing sensitive response bands concept for image processing and mapping of basalts q Karim W. Abdelmalik Geology Department, Faculty of Science, Ain Shams University, Cairo, Egypt

a r t i c l e

i n f o

Article history: Received 30 August 2018 Revised 21 February 2019 Accepted 22 April 2019 Available online xxxx Keywords: Landsat 8 Band algebra PCA ICA Band ratios

a b s t r a c t The spectral signature curve behavior of basalt and its relationship with the surface reflectance (SR) values of Landsat 8 OLI data, (Operational Land Imager), is considered as the key factor for determination of the sensitive response bands for basalt interaction to be used in both Object-Oriented Principal Components (PCA) & Independent Component (ICA) analyses. Also, it is extremely beneficial in develop mathematical formulas which express the best band ratios for the precise illustration and mapping of basalts. The integration of those sensitive bands and the three mentioned methods are the main objectives of the present investigation. This should permit to create the basalt spatial distribution map for a given area. These methods were applied in west central Sinai, Egypt. The measured spectral signature reflectance curve of the basalt was resampled to meet the spectral characteristics of Landsat 8 bands, both curves were carefully examined to determine the most significant response bands for basalts; which were found to be bands 4, 5, 6 and 7. The examined spectral behavior led to three significant band ratios. Moreover, two false color composite (PC1, PC2, PC4) and (IC4, IC3, IC1) in RGB, were determined by execution and examination of the PCA and ICA images, respectively of the study area. ERDAS Imagine (2010) and ArcGIS 10.3 packages were used for digital/mathematical processing steps and to apply the resulted models on the study area. Ó 2019 National Authority for Remote Sensing and Space Sciences. Production and hosting by Elsevier B. V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/).

1. Introduction Geological mapping, in all scales, is one of the main targets for geologists. However, the traditional methods of geologic field mapping are expensive and time consuming, when compared with remote sensing techniques and methods, particularly concerning large areas. (Abdelmalik, 2018) Remote sensing technology has rapid growing reaching superior advanced levels that are convenient for geological applications as structural and lithological mapping as well as economic minerals exploration and mapping (Abrams et al., 1983; Abrams and Hook, 1995; Sabins, 1997; Rowan and Mars, 2003; Rowan et al., 2003; Mars and Rowan, 2006; Gad and Kusky, 2006, 2007; Amer et al., 2010).

Peer review under responsibility of National Authority for Remote Sensing and Space Sciences. E-mail addresses: [email protected], [email protected]. edu.eg

Numerous, mathematical/statistical, algorithms and processes have been developed and used by many authors to emphasize and enhance the geologic features/targets in the remotely sensed data. The widely used and most effective methods for geological applications (e.g. lithological mapping and mineral exploration) are band ratio (BR) and principal component analysis (PCA). The real power of the remote sensing techniques arises through integration of the spectral signature curves of the investigated targets with the remotely sensed data, that fact is a result based on the information contained in the spectral signature curves of each material which is sensitive for both physical and chemical characteristics of any object. The focal objectives of this article aim to: (1) Study and analyze the spectral curves of basalts to emphasize its interaction behavior with the electromagnetic spectrum as well as to discover the intrarelation behavior among various recorded wave lengths, (2) Reveal the most sensitive bands for the basalts. and (3) Generate the most effective band ratios, object-oriented principal component and independent component analyses based on the selected bands, for emphasizing and mapping the basalt targets.

https://doi.org/10.1016/j.ejrs.2019.04.004 1110-9823/Ó 2019 National Authority for Remote Sensing and Space Sciences. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

2

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

2. Geology of the area The inactive continental rift of the Gulf of Suez (Suez rift) lies at the northwestern terminus of the Red Sea active rift and between the north Eastern Desert of Egypt and south Sinai (Fig. 1). The Suez rift is an elongated basin of thin lithosphere bounded by massifs of relatively thicker crust (Steckler et al., 1988). Opening of this rift started at about Oligocene or Oligo-Miocene time (e.g. Robson, 1971; Angelier, 1985; Evans, 1988; Lyberis, 1988). The Oligo-Miocene basaltic volcanism and the Oligocene Red Beds in Hammam Faraun block are two evidences that characterize the beginning of the Suez rifting. Basaltic dikes, sills and flows are exposed at both margins of the rift and have been encountered in a number of exploratory wells in the rift trough (Patton et al., 1994). The best exposures of these basaltic sheets exist in the western part of the Hammam Faraun block. Stratigraphic relationship dates these sheets to post-late Eocene and pre-earlist Miocene (Moustafa and Abdeen, 1992; Patton et al., 1994; Moustafa, 1996) and their absolute age ranges between 22 and 26 Ma (Steen, 1984, Moussa, 1987, Meneisy, 1990). These basaltic sheets have subalkaline to alkaline affinities and were intruded in a tensional regime (Moussa, 1987). Dikes orientations are mainly northwest to north-northwest; parallel to the rift trend; although other less common orientations which contain east–west, north-northeast, northeast, and north– south also exist (Patton et al., 1994). The western part of Hammam Faraun block was extruded by several basaltic sheets which have different trends. So, this part was chosen to be a focused target of the current article. The exposed stratigraphic units in the western part of the Hammam Faraun block are divided into three tectono-stratigraphic

Fig. 1. General map of Sinai showing the distribution of major basaltic dikes and other intrusive bodies associated with the Oligo-Miocene magmatism (Modified after El-Bialy et al., 2017).

sequences with respect to the Oligo-Miocene Suez rift. These sequences are pre-rift, syn-rift and post-rift (Fig. 2), The post-rift sequence includes the Pliocene-Quaternary sediments, which fill the present-day wadis and alluvial plains. Generally, the pre-rift sequence contains a wide variety of rocks of Precambrian-Eocene age. However, in the study area, the pre-rift rocks belong to the Upper Cretaceous mixed facies and Upper Cretaceous-Eocene carbonate facies. The Upper Cretaceous clastic and carbonate beds of the Matulla and Duwi Formations in the area just east of Abu Zenima city. The measured thicknesses of these two formations attain 112 m and 41 m, respectively. The Sudr Chalk; 189 m thick and Lower Eocene Thebes Formation; 307 m thick. The Paleocene Esna Shale (32 thick) exists between these two formations. The Middle Eocene Darat and Tanka Formations overlie the Thebes Formation in other parts of the study area. The syn-rift sequence contains Oligocene and Miocene units, which are separated by a regional unconformity. The Oligocene Red Beds (Abu Zenima Formation) and the Oligo-Miocene basalts represent the basal syn-rift units (Fig. 2). The Red Beds (54 m thick) consist of ferruginous sandstone, mudstone, and shale. This forma-

Fig. 2. Stratigraphy of the investigated area.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

tion is exposed only in the area around the mouth of Wadi Tayiba as well as in Wadi Nukhul and Gebel Tanka. The Miocene Nukhul and Rudies Formations are exposed only in the Wadi Tayiba-Gebel Tanka area as well as in Wadi Nukhul. The shallow-marine sandstone and conglomerate of the Lower Miocene Nukhul Formation exist under the deeper marine marl, shale, limestone, and sandstone of the Rudeis Formation. 3. Methodology In the current investigation, both Landsat 8 (Operational Land Imager; OLI) and the measured spectral signature curve of the studied basalts samples, are used for determining the most sensitive bands for the target basalts. The selected bands used for deriving and executing the most successful ratios, object-oriented PCA and ICA to produce clear and rather more interpretable images to precise differentiation and discrimination of the basalts in the investigated area. The Landsat 8 is an American earth observation satellite, which was launched finally on February 2013. It is the latest of Landsat series. It is ongoing partnership between USGS and NASA. The mission of Landsat 8 is mainly targeting to offer regular, high quality of visible, infrared and thermal images so it has two sensors, OLI and TIR and cover the spectral ranges visible and near infrared, VNIR (0.435–0.673 mm), short wave infrared, SWIR (1.56–2.29 mm) and thermal range (10.6–12.5 mm) with spatial resolution 30 m for VNIR/SWIR, 100 m (resampled to 30 m) for thermal range and 15 m for panchromatic band (Table. 1). The data used in the present study is Landsat 8 (Path 175/Row 40) obtained from the home site of Earth Explorer (Earth explorer.usgs.gov) and acquired on 28/8/2013, the data is geometrically corrected and georeferenced to UTM projection (Zone 36 N), WGS-84 datum and ellipsoid. 3.1. Petrographic description The petrographic description of the collected samples was performed in transmitted light using PriorLux Pol Polarizing Microscope. In addition to the XRD examination were executed using a PANaLytical X-Ray Diffraction equipment model X’Pert PRO with Monochromator, Cu-radiation (k = 1.542 Å) at 45 K.V., 35 M.A. and scanning speed 0.03°/s. the diffractograms and relative intensities were compared with ICDD files.

The core of the effective processing and/or interpretation chain is understanding the interaction behavior between electromagnetic spectrum (EMS) and the targets. Many authors worked on

Table 1 Spectral characteristics and spatial resolutions of Landsat8 bands.

VNIR

SWIR TIR

measuring of the reflectance of rocks and minerals (e.g. Korb et al., 1996; Salisbury et al.,1991; Kokaly et al., 2017; Howari et al., 2018) to create and build spectral libraries such as USGS (https://crustal.usgs.gov/speclab) and JHU (https://speclib.jpl.nasa.gov/) libraries. Those spectral libraries contain a pool of spectral signature curves of many rocks. The integration of both measured and observed spectral signature curves with the remotely sensed imagery have been used by many authors in order to identify and map the different rock units (e.g. Ducart et al., 2006; Harris et al., 2010; Madani, 2011, 2014, 2015; Nair et al., 2014; Wang et al., 2014; Boesche et al., 2015; Emam et al., 2016; Ahmad et al., 2016; Abdelmalik and AbdAllah, 2018; Babu et al., 2015; Chenlong et al., 2018). Several authors provide the efficiency of ASD fieldspec measured data either inside or indoor (e.g. Madani, 2011, 2014, 2015; Madani and Harbi, 2012; Wang et al., 2014) Madani (2015) studied the spectral data of olivine basalt in wadi Natash area (Southern Eastern Desert, Egypt) and defined five absorption characterized features of olivine basalts. In the present study, the ASD high resolution Fieldspec (Table 2) was utilized for measuring the spectral reflectance curves of basalt samples. The obtained reflectance profile was subjected to resampling algorithm to match the spectral characteristics and specifications of the OLI sensor of Landsat 8. Both the original and the resampled curves were carefully studied and examined to select the most sensitive bands for basalts to be used for creation and development appropriate and optimal band ratios and object oriented PCA in addition to ICA processes to raise the best and precise discrimination/mapping of basalts. 3.3. Remotely sensed images pre-processing and processing 3.3.1. Atmospheric correction The recorded signal reflected from the earth’s surface is considered as an outcome of two components; the actual reflectance from target component and noise one resulted from the atmospheric effect. Because of this fact, the atmospheric correction is considered as one of an essential step of pre-processing phase (e.g. Liang et al., 2001; Lillesand et al., 2004; Chander et al., 2009; Tyagi and Bhosle, 2011; Abdelmalik, 2018; Abdelmalik and Abd-Allah, 2018).

Table 2 Specification of the ASD High Res. FeildSpec used in the study.

3.2. Spectral signature curves study

Band

Wavelength Range (mm)

Spatial Resolution (m)

Remarks

1

0.43–0.45

30

2 3 4 5

0.45–0.51 0.53–0.59 0.64–0.67 0.85–0.88

Ultra-Blue (coastal/aerosol) Blue Green Red NIR

8 9 6 7

0.50–0.68 1.36–1.38 1.57–1.65 2.11–2.29

15 30

Panchromatic Cirrus SWIR1 SWIR2

10 11

10.60–11.19 11.50–12.51

100

TIR1 TIR2

3

Description

specification

Spectral Range Spectral Resolution

350–2500 nm 3 nm @ 700 nm 8 nm @ 1400/2100 nm 1.4 nm @ 350–1000 nm 1.1 nm @ 1001–2500 nm 100 ms VNIR 0.02%, SWIR 1 & 2 0.01%

Spectral sampling (bandwidth) Scanning Time Stray light specification Wavelength reproducibility Wavelength accuracy Maximum radiance Channels Detectors

Noise Equivalent Radiance (NEdL)

0.1 nm 0.5 nm VNIR 2X Solar, SWIR 10X Solar 2151 VNIR detector (350–1000 nm): 512 element silicon array SWIR 1 detector (1001–1800 nm): Graded Index InGaAs Photodiode, Two Stage TE Cooled SWIR 2 detector (1801–2500 nm): Graded Index InGaAs Photodiode, Two Stage TE Cooled VNIR 1.0
109 W/cm2/nm/sr @700 nm SWIR 1 1.4
109 W/cm2/nm/sr @ 1400 nm SWIR 2 2.2
109 W/cm2/nm/sr @ 2100 nm

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

4

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

In the present study, the radiometric calibration tool provided in ENVI 5.1 was used for calibrating the Landsat 8, OLI, image data (V, NIR and SWIR regions) to reflectance. This process calculates Top-Of-Atmosphere reflectance (TOA) and executed based on some factors such as acquisition time; sun elevation; gains; offsets and solar irradiance, finally the surface reflectance (SR) was calculated. 3.3.2. Band ratio and image algebra Generally, digital enhancement of any image is a process targeting to make the target image clearer and easier to interpret for a particular application (Faust, 1989). There are numerous enhancement techniques based on the objectives, one of the most important enhancement methods for lithological mapping is image algebra or band ratio in which a mathematical formula created and applied on the matrix of digital image to emphasize the inves-

tigated target and eliminate the shadow effects caused by topography. This method was used and applied by many authors to distinguish, differentiate as well as mapping variant rock units (e.g. Sultan and Arvidson, 1986; Xu et al., 2004; Gomez et al., 2005; Ninomiya et al., 2005; Gad and Kusky, 2006; Amer et al., 2010; Zoheir and Emam, 2012; Matar and Bamousa, 2013; Ding et al., 2014; Elmagd et al., 2015; Rajendran and Nasir, 2014; Lamri et al., 2016; Guha and Vinod Kumar, 2016; Abdelmalik and Abd-Allah, 2018). Moore et al. (2007) had utilized Landsat7 data to perform a quantitative study of basaltic rock outcrops. Moreover, Drury (1993) and Darning (1998) reported some succeed ratios for differentiating and identifying lithologies and rock types (e.g. 3/1, 3/5, 3/7, 5/1, 5/4, 3/7, and 5/7). It is also found that the false color ratio image 5/4, 5/1 and 3/7 in RGB, is considered as the best result for the basalt rocks discrimination (with reddish color). Corumluoglu

Fig. 3. Graphical representation of eigenvector matrix of Landsat 8 (OLI) bands. Tables showing the eigenvector statistics for the four selected bands (4, 5, 6 and 7). The bands were selected in order to identify the high spectral responses of the basalt rocks. (a) eigenvector matrix resulted from PCA (b) eigenvector matrix resulted from ICA.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

5

Fig. 4. Representative photomicrographs of the basaltic dikes from the study area (a) Augite crystal enclosed by prismatic plagioclase laths forming doleritic texture, X40, XPL; (b) Cracked olivine phenocryst, X40, XPL; (c) A secondary calcite veinlet leaving very small relicts of preexisting minerals, X40, PPL; (d) Subophitic texture of augite crystal partially enclosing lath-shaped euhedral crystals of plagioclase in an irregular pattern, X40, XPL; (e) Sericitization along the cleavage of a plagioclase crystal, X40, XPL; (f) Altered pyroxene laths and seritized plagioclase crystals, X40, XPL.

et al. (2015) studied the best methods for determination of Kula basalts, Turkey using Landsat 7 (ETM+). They found that, the best band ratio composite image for mapping the basalts is 3/1, 4/5, and 3/2 in RGB, which represents the basalts in dark blue color. Madani (2015) assigned 3/2, 8/1 and 8/5 ASTER band ratios for identification of olivine basalt. In this current study, the determined behavior as well as the sensitive bands of basalts resulted from the deep and careful study for both, original spectral signature curves and resampled one, were used for illation and development the best and most appropriate mathematical formula to enhance and emphasize the basalts in the study area.

3.3.3. Object-Oriented principal component analysis (PCA) Generally, principal component analysis (PCA) is one of the most powerful and effective image enhancement methods specially for geological mapping (e.g. Lithological and mineralogical applications), this is as a result of the output of noncorrelated PC-bands as well as isolation and marginalization of noise. The computing of principal components transformation (PCT) executed by a linear transformation of the data using eigenvectors and eigenvalues which mathematically derived from the covariance matrix (Eq. (1)) then the original data file values transformed into the principal component values (Eq (2)) (Jensen, 1996; Faust, 1989, Gonzalez and Wintz, 1977)

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

6

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

Fig. 5. Representative X-Ray Diffractogram of basalt from study area.

Fig. 6. (a) Spectral signature curve of abundant minerals in investigated Basalt, available via USGS spectral library Version 7 (https://crustal.usgs.gov/speclab), (b) Spectral profile of altered basalt samples measured by ASD Fieldspec Hi-Res in laboratory (c) the resampled spectral curve of basalt using the spectral characteristics of Landsat 8, the gray vertical lines illustrate OLI bands.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

7

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

3

2

m1 0 0    0 6 0 m 0  0 7 2 7 6 7V ¼ Ee Cov ðEe ÞT 6 4    5 0

0

0



ð1Þ

mn

    PP  Y i Y i i¼1 X i  X i ffirffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ; 0  j Rxy j  1 Rxy ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2 PP  PP  X  X Y  Y i i i i i¼1 i¼1

ð3Þ

where:

where: Cov = the matrix of covariance Ee = eigenvectors matrix T = function of transposition V = The eigenvalues diagonal matrix (nondiagonal elements are zeros)

Pe ¼

Xn k¼1

ðdk Eke Þ

ð2Þ

where: e = the first and second principal component number Pe = the output value of principal component of (e) number k = an input band n = bands numbers (total) dk = the value (input) in band k Eke = the element of eigenvector matrix at k and e (row, column) However, applying the PCA on the sensitive bands of the target material make the method more efficient and powerful (Crósta et al., 2003; Gaber et al., 2010, 2015). According to Corumluoglu et al. (2015) The composite image of PC4, PC3, and PC2 of Landsat’s PCA six reflected bands (1, 2, 3, 4, 5 and 7) was the best to illustrate the basalts. Moreover, he performed the PCA to two sets of four-band subsets (1, 3, 4 and 5) and (1, 4, 5 and 7) then used PC4 resulted from both PCA images to map the basalts. In the present study, spectral subset of four bands of Landsat 8 data (bands 4, 5, 6 and 7), the high response bands concluded through studding the spectral curves, original and resampled ones, were subjected to PCA process. The selected PCs of the outcome PC image were based on the intra-relations of eigenvector (Fig. 3) particularly the high eigenvector anomaly even it was in opposite directions (+ve and ve) to produce more interpretable image focused on the investigated target. 3.3.4. Independent component analysis (ICA) One of the superior targets for image processing and interpretation for geological applications is feature identification and extraction. Independent component analysis (ICA) is a high order statistical feature extraction method, it exploits the higher order statistical-characteristics of the remotely sensed data. Therefore, the output ICA components not only uncorrelated, like PCA components, but also independent; and each band of the independent component (IC) comprise information of specified feature which is corresponding to that in the original data. (e.g. Shah, 2003; Common, 1994). In the present study, the ICA component ordering preformed based on the correlation coefficient which is a measure of similarity between two images and obtained with Eq. (3).

(Rxy) Correlation coefficient between two images X and Y P The total number of image pixels 4. Results and discussion 4.1. Petrographic description Petrographically the studied dikes are identified as basalt and the coarser variety basaltic andesite (Fig. 4). Phenocrysts in these rocks are euhedral plagioclase, which are partially altered to clay minerals and sericite along the mineral cleavage planes (Fig. 4e and f), diopside-rich augite which is occasionally replaced by hornblende and few crystals of highly cracked olivine. The matrix is microgranular, holocrystalline and composed mainly of plagioclase, diopside, augite and opaque minerals. Ophitic, subophitic and doleritic textures are common (Fig. 4a and d). The studied dikes are slightly altered with variable degree. X-Ray Diffractograms of the collected samples confirm that the mineralogical composition of basalt is dominated by diopside, augite, anorthite, analcime, tschernichite and lepidocrocite (Fig. 5). 4.2. Spectral signature curve characteristics The careful study of the basalt’s spectral curve and the deep examination of the resampled one (Fig. 6) led to better understanding the reflectance behavior of basalts as well as determining the most reflectance response/sensitive bands for it. The measured spectral profile for the representative investigated basalt samples (Fig. 6) shows five absorption anomalies. In which the first one is represented at the range 0.35–0.45 mm, this is correlated to the presence of lepidocrocite and hornblende (very low reflectance) in addition to augite and olivine (low reflectance). The second absorption range located around 1 mm as a result of Fe-Oxides presence and to absorption features of olivine, augite, diopside and analcime (Fig. 6). The third absorption anomaly is around 1.4 mm due to the major absorption of analcime (at  1.45 mm), the wide absorption of olivine (0.9–1.45 mm) and the minor absorption of both anorthite and hornblende (Fig. 6). The fourth main absorption anomaly showed in the measured spectral profile within the range 1.95 mm which is attributed to the major absorption anomaly of analcime associated with the minor absorption of calcite, lepidocrocite and anorthite (Fig. 6). Finally, the highly absorption feature of both carbonates and hornblende accompanying with minor absorption feature of anorthite are witnessed around 2.35 mm. The resampled spectral profile of the altered basalts (Fig. 6) together with the previously mentioned observations and interpretation encapsulate the distinguished behavior of the basalt with respect to Landsat 8 bands. This behavior is evaluated as low reflectance behavior at bands 1 and 2 (0.43–0.45 mm and 0.45–

Table 3 Spectral characteristics of the selected sensitive bands for basalts.

Sensitive bands Characteristics

Band

Wavelength Range (mm)

Wavelength Midpoint (mm)

Reflectance Behavior

Band Name

4 5 6 7

0.63–0.67 0.85–0.88 1.57–1.65 2.11–2.29

0.650 0.865 1.610 2.200

High Low High Low

Red NIR SWIR1 SWIR2

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

8

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

0.51 mm, respectively) and sharp increase of reflectance at ranges of bands 3 and 4 (0.53–0.59 mm and 0.63–0.67 mm, respectively) with decrease tendency around band 5 (0.85–0.88 mm) and shows a high reflectance response at the range of band 6 (1.57–1.65 mm) then decreasing reflectance tendency within range of band 7 (2.11–2.29 mm). four bands were selected as the most sensitive ones for basalts (Table 3).

4.3. Band ratio Based on the four-selected sensitive and high respond bands which resulted and obtained from the examined spectral behavior as well as the resampled spectral curve, three band ratios were developed for clear illustrations and mapping of basalts in studied region.

Fig. 7. (a) Landsat 8 7, 5, 3 (R, G, B) of study area illustrates the locality of the two examined areas (I and II). (b), (c) and (d) show the results of the three developed band ratio formulas for both tested areas (I and II), basalts separated with bright tone. (b) resulted image for (band2-band7)/(band4 + band3) (Eq. (4)), (c) resulted image for (band3band7)/(SQRT(band4)) (Eq. (5)), and (d) resulted image for [(band3-band7)/CONVOLVE (band4)] (Eq. (6)).

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

9

4.4. PCA images

The successful band mathematical formulas are:

ðband2  band7Þ=ðband4 þ band3Þ

ð4Þ

ffiffiffiffiffiffiffiffiffiffiffiffiffiffi p 2 ðband3  band7Þ=ð band4Þ

ð5Þ

  1 1 1     ðband3  band7Þ= 1 3 1 ðband4Þ   1 1 1

ð6Þ

All developed mathematical band ratios were visually examined. All resulted ratio images emphasize and illustrate the basalts in excellent way with very bright tone (Fig. 7). However, the most illustrative and precise one is that represented in Eq. (5) due to the extreme difference between the numerator and denominator of the mathematical formula (Fig. 7c)

Both obtained eigenvector values and the original bands information percentage loaded in each PC (Fig. 3) show that PC2 and PC4 illustrate an opposite anomalies of eigenvector values loads from bands (4, 5 as -ve and 6, 7 as +ve) and (4, 6 as -ve and 5, 7 as +ve) respectively, which exhibit the spectral signature from basalts (Table 3). The selected PC bands according to their maximum data loaded as well as the sensitive target differentiation were PC2 and PC4 as previously mentioned, and The band composite which proposed for the optimal illustration of the obtained anomalies (basalts) was PC2 (as green) and PC4 (as blue); moreover, PC1 (as red) was selected visually as it shows the best third band in the combination. The resulted false color composite images PC1, PC2, PC4 in RGB was extremely successful in mapping all basalt dikes in the study area, which are emphasized with dark olive-green color (Fig. 8). This is mainly because PC2 was emphasizing the effect of band 6 (high reflectance behavior) in +ve direction and band 5 (low reflec-

Fig. 8. (a) PCA False color composite PC1, PC2, PC4 (R, G, B) of study area illustrates the locality of the two examined areas. (b) and (c) magnifying of the examined areas (refer to (a) for locations), the basalt dikes are shown in dark olive-green color.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

10

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

tance behavior) in ve direction and that lead to high brightness behavior for basalts. 4.5. ICA images According to the selected sensitive bands for the basalt (Table 3), a four-bands spectral subset image was subjected to the ICA process. The resulted bands of ICA image were loaded with pixels anomaly according to unique spectral behavior features and were earmark for RGB color combination to obtain a more interpretable image for the investigated region (Fig. 9). The resulted image is derived from RGB color combination of IC4 (band 4 of Landsat 8) as red, IC3 (band 7 of Landsat 8) as green

and IC1 (band 5 of Landsat 8) as blue as shown in graphical representation of eigenvector matrix of the ICA (Fig. 3). The selected IC components loaded by statistical factors which are independent from each other, moreover, these IC data layers are the most sensitive for spectral interaction of the basalts. The IC band combination was very successful in extracting and mapping the basalt dikes in the study area, which were distinguished with dark green color (Fig. 9). From the three sections mentioned above, the comparison among the three utilized studied techniques exhibits a great engagement for all resulted outputs. Although being quite similar for basalts recognition within the investigated area, the better discrimination is revealed band ratio technique which yields best

Fig. 9. (a) ICA False color composite 4, 3, 1 (R, G, B) of the study area illustrates the locality of the two examined areas. (b) and (c) magnifying of the examined areas (refer to (a) for locations), the basalt dikes are shown in dark green color.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

interpreted results rather than the others. It improves the band algebra results with direct usage of the most sensitive response predefined bands in the second developed mathematical equation (Eq. (5), Fig. 7c). 5. Conclusion The remotely sensed satellite-based approach used in the present investigation to enhance, extract and precis mapping of basalt dikes has been proved to be an extremely successful tool. It could be comprehensively used for basalts mapping in other inaccessible regions. The integration of the spectral signature curves with all used enhancement methods and processes showed fruitful results. The newly developed Landsat 8 (OLI) band ratios and both Object-Oriented Principal Component (PCA) as well as Independent Component (ICA) Analyses images were extremely useful in extracting the basalt dikes, the best developed band ratio in the present study is (band3-band7)/(SQRT(band4)). The successful false color composites – in RGB – for mapping the basalt dikes were PC1, PC2, PC4 and IC4, IC3, IC1 respectively. The best and most pricing way for mapping the basalts among all the tested models was revealed in the band ratio method. Acknowledgments The author would like to thank the anonymous reviewers for their valuable comments and suggestions to improve the quality of the paper as well as National Authority for Remote Sensing and Space Sciences (NARSS) for its influenced assistance in measuring the spectral profile of investigated samples. Also deep thanks for Prof. Yehia H. Dawood of Ain Shams University for his assistance in petrographic and XRD investigation. Conflict of interest There is no conflict of Interest.

References Abdelmalik, K.W., 2018. Role of statistical remote sensing for inland water quality parameters prediction. Egypt. J. Remote Sens. Sp. Sci. 21 (2), 193–200. https:// doi.org/10.1016/j.ejrs.2016.12.002. Abdelmalik, K.W., Abd-Allah, A.M.A., 2018. Integration of remote sensing technique and field data in geologic mapping of an ophiolitic suture zone in western Arabian Shield. J. Afri. Earth Sci. 146, 180–190. https://doi.org/10.1016/ j.jafrearsci.2017.10.006. Abrams, M.J., Brown, D., Lepley, L., Sadowski, R., 1983. Remote sensing for porphyry copper deposits in southern Arizona. Econom. Geol. 78, 591–604. Abrams, M.J., Hook, S.J., 1995. Simulated ASTER data for geologic studies. IEEE Trans. Geosci. Remote Sens. 33, 692–699. Ahmad, L., Shah, M.T., Khan, S.D., 2016. Reflectance spectroscopy and remote sensing data for finding sulfide-bearing alteration zones and mapping geology in Gilgit-Baltistan. Pak. Earth Sci. Inf. 9, 113–121. Amer, R., Kusky, T.M., Ghulam, A., 2010. Lithological mapping in the Central Eastern Desert of Egypt using ASTER data. J. Afri. Earth Sci. 56, 75–82. Angelier, J., 1985. Extension and rifting: the Zeit region, Gulf of Suez. J. Struct. Geol. 7, 605–612. Babu, P.S., Majumdar, T.J., Bhattacharya, A.K., 2015. Study of spectral signatures for exploration of bauxite ore deposits in Panchpatmali, India. Geocarto Int. 30, 545–559. Boesche, N., Rogass, C., Lubitz, C., Brell, M., Herrmann, S., Mielke, C., Tonn, S., Appelt, O., Altenberger, U., Kaufmann, H., 2015. Hyperspectral REE (rare earth element) mapping of outcrops—Applications for Neodymium detection. Remote Sens. 7, 5160–5186. Chander, G., Markham, B.L., Helder, D.L., 2009. Summary of current radiometric calibration coefficients for Landsat MSS, TM, ETM+, and EO-1 ALI sensors. Remote Sens. Environ. 113 (5), 893–903. Shi Chenglong, DingXiaozhong, Liu Yanxue, Zhou Xiaodong. 2018. Reflectance Spectral Features and Significant Minerals in Kaishantun Ophiolite Suite, Jilin Province, NE China. Minerals, https://doi.org/10.3390/min8030100

11

Common, P., 1994. Independent component analysis, a new concept? Signal Processing 36, 287–314. Corumluoglu, O., Vural, A., Asri, I., 2015. Determination of Kula basalts (geosite) in Turkey using remote sensing techniques. Arab. J. Geosci. 8, 10105–10117. https://doi.org/10.1007/s12517-015-1914-4. Crósta, A.P., Souza Filho, C.R., Azevedo, F., Brodie, C., 2003. Targeting key alteration minerals in epithermal deposits in Patagonia, Argentina, using ASTER imagery and principal component analysis. Int. J. Remote Sens. 24, 4233–4240. Darning, W.P., 1998. Affiliated Research Center, Integrated Use of Remote Sensing and GIS for Mineral Exploration. Final Report, 3–4. Ding, C., Liu, X., Liu, W., Liu, M., Li, Y., 2014. Mafic and ultramafic and quartz-rich rock indices deduced from ASTER thermal infrared data using a linear approximation to the Planck function. Ore Geol. Rev. 60, 161–173. Drury, S.A., 1993. Image interpretation in Geology. Chapman and Hall, London, p. 304. Ducart, D.F., Crosta, A.P., Filho, C.R.S., Coniglio, J., 2006. Alteration mineralogy at the Cerro La Mina epithermal prospect, Patagonia, Argentina. field mapping, shortwave infrared spectroscopy, and ASTER images. Econ. Geol. 101, 981–996. El-Bialy, M.Z., Khalifa, I.H., Omar, M.M., 2017. Continental intraplate volcanism in the Sinai subplate: the Oligo-Miocene basalts of the Gulf of Suez rift. J. Afri Earth Sci. https://doi.org/10.1016/j.jafrearsci.2017.03.004. El-magd, I.A., Mohy, H., Basta, F., 2015. Application of remote sensing for gold exploration in the Fawakhir area, central Eastern desert of egypt. Arab. J. Geosci. 8, 3523–3536. https://doi.org/10.1007/s12517-014-1429-4. Emam, A., Zoheir, B., Johnson, P., 2016. ASTER-based mapping of ophiolitic rocks: examples from the Allaqi-Heiani suture, SE Egypt. Int. Geol. Rev. 58, 525–539 https://doi:10.1080/00206814.2015.1094382. Evans, A.L., 1988. Neogene tectonic and stratigraphic events in the Gulf of Suez rift area. Egypt. Tectonophysics 153, 235–247. Faust, N.L., 1989. Image enhancement. In: Kent, A., Williams, J.G. (Eds.), Supplement 5 of Encyclopedia of Computer Science and Technology, vol. 20. Marcel Dekker Inc, New York. Gabr, S., Hassan, S., Sadek, F., 2015. Prospecting for new gold-bearing alteration zones at El-Hoteib area, South Eastern Desert, Egypt, using remote sensing data analysis. Ore Geol. Rev. 71, 1–13. Gabr, S., Ghulam, A., Kusky, T., 2010. Detecting areas of high-potential gold mineralization using ASTER data. Ore Geol. Rev. 38, 59–69. Gad, S., Kusky, T.M., 2006. Lithological mapping in the Eastern Desert of Egypt, the Barramiya area, using Landsat Thematic Mapper (TM). J. Afri. Earth Sci. 44, 196– 202. Gad, S., Kusky, T.M., 2007. ASTER spectral ratioing for lithological mapping in the Arabian-Nubian shield, the Neoproterozoic Wadi Kid area, Sinai. Egypt. Gondwana Res. 11 (3), 326–335. Gomez, C., Delacourt, C., Allemand, P., Ledru, P., Wackerle, R., 2005. Using ASTER remote sensing data set for geological mapping, in Namibia. Phys. Chem. Earth. 30, 97–108. Gonzalez, R.C., Wintz, P., 1977. Digital Image Processing. Addison-Wesley Publishing Company, Reading, Massachusetts. Guha, A., Vinod Kumar, K., 2016. New ASTER derived thermal indices to delineate mineralogy of different granitoids of an Archaean Craton and analysis of their potentials with reference to Ninomiya’s indices for delineating quartz and mafic minerals of granitoidsdan analysis in Dharwar Craton. India. Ore Geol. Rev. 74, 76–87. Harris, J.R., Mcgregor, R., Budkewitsch, P., 2010. Geological analysis of hyperspectral data over southwest Baffin Island: Methods for producing spectral maps that relate to variations in surface lithologies. Can. J. Remote Sens. 36, 412–435. Howari, F.M., Acbas, G., Nazzal, Y., AlAydaroos, F., 2018. Hapke-based computational method to enable unmixing of hyperspectral data of common salts. Chem. Central J. https://doi.org/10.1186/s13065-018-0460-z. Jensen, J.R., 1996. Introductory Digital Image Processing: A Remote Sensing Perspective. Prentice-Hall Englewood Cliffs, New Jersey, pp. 197–256. Kokaly, R.F., Clark, R.N., Swayze, G.A., Livo, K.E., Hoefen, T.M., Pearson, N.C., Wise, R. A., Benzel, W.M., Lowers, H.A., Driscoll, R.L., and Klein, A.J., 2017. USGS Spectral Library Version 7: U.S. Geological Survey Data Series 1035, pp. 61, https://doi. org/10.3133/ds1035. Korb, A.R., Dybwad, P., Wadsworth, W., Salisbury, J.W., 1996. Portable FTIR spectrometer for field measurements of radiance and emissivity. Appl. Opt. 35, 1679–1692. Lamri, T., Djemaï, S., Hamoudi, M., Zoheir, B.A., Bendaoud, A., KhadidjaOuzegane, K., Massinissa, A., 2016. Satellite imagery and airborne geophysics for geologic mapping of the Edembo area, Eastern Hoggar (Algerian Sahara). J. Afr. Earth Sci. 115, 143–158. Liang, S., Fang, H., Chen, M., 2001. Atmospheric correction of Landsat ETM+ land surface imagery-part I: methods. IEEE Trans. Geosci. Remote Sens. 39 (11), 2490–2498. Lillesand, T.M., Kiefer, R.W., Chipman, J.W., 2004. Remote Sensing and Image Interpretation. John Wiley and Sons, New York. Lyberis, N., 1988. Tectonic evolution of the Gulf of Suez and the Gulf of Aqaba. Tectonophysics 153, 209–220. Madani, A., 2014. Assessment and evaluation of band ratios brovey and HSV techniques for lithologic discrimination and mapping using Landsat ETM+ and SPOT-5 data. Int. J. Geosci. 5, 5–11. Madani, A., 2015. Spectroscopy of olivine basalts using FieldSpec and ASTER data: A case study from Wadi Natash volcanic field, south Eastern Desert. Egypt. J. Earth Syst. Sci. 124 (7), 1475–1486.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004

12

K.W. Abdelmalik / Egypt. J. Remote Sensing Space Sci. xxx (xxxx) xxx

Madani, A., 2011. Spectral properties of carbonatized ultramafic mantle xenoliths and their host olivine basalts, Jabal Al Maqtal basin, South Eastern Desert, Egypt, using ASD Field Specspectroradiometer. Egypt. J. Remote Sens. Sp. Sci. 14, 41– 48. Madani, A., Harbi, H., 2012. Spectroscopy of the mineralized tonalite–diorite intrusions, Bulghah gold mine area, Saudi Arabia: Effects of opaques and alteration products on FieldSpec data. Ore Geol. Rev. 44, 148–157. Mars, J.C., Rowan, L.C., 2006. Regional mapping of phyllic- and argillic-altered rocks in the Zagros magmatic arc, Iran, using Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data and logical operator algorithms. Geosphere 2 (3), 161–186. Matar, S.S., Bamousa, A.O., 2013. Integration of the ASTER thermal infra-red bands imageries with geological map of Jabal Al Hasir area, Asir Terrane, the Arabian Shield. J. Taibah Univ. Sci. 7 (1), 1–7. Meneisy, M., 1990. Vulcancity. In: Said, R. (Ed.), The Geology of Egypt. Balkema Rotterdam, pp. 157–174. Moore, C.A., Hoffmann, G.A., Glenn, N.A., 2007. Quantifying basaltic rock outcrops in NRCS soil map units using Landsat-5 Data. Soil Surv. Horiz. 48, 59–63. Moussa H.E., 1987. Geologic studies and genetic correlation of basaltic rocks in west central Sinai. Ph.D. Thesis, Ain Shams University, Faculty of Science, Egypt, pp. 308. Moustafa, A.R., Abdeen, M.M., 1992. Structural setting of the Hammam Faraun block, Eastern side of the Suez rift. J. Univ. Kuwait (Science) 19, 291–309. Moustafa, A.R., 1996. Structural setting and tectonic evolution of the northern Hammam Faraun block (Wadi Wasit-Wadi Wardan area), eastern side of the Suez rift. Kuwait J. Sci. Eng. 23, 105–132. Nair, A.M., Mathew, G., 2014. Effect of bulk chemistry in the spectral variability of igneous rocks in VIS-NIR region: implications to remote compositional mapping. Int. J. Appl. Earth Obs. Geoinform. 30, 227–237. Ninomiya, Y., Fu, B., Cudhy, T.J., 2005. Detecting lithology with advanced Spaceborne Thermal Emission and Refection Radiometer (ASTER) multispectral thermal infrared ‘‘radiance-at-senseor” data. Remote Sens. Environ. 99, 127–135. Patton, T.L., Moustafa, A.R., Nelson, R.A., Abdine, S.A., 1994. Tectonic evolution and structural setting of the Suez rift. In: Landon, S.M. (Ed.), Interior Rift Basins. American Association of Petroleum Geologists Memoir, pp. 9–55.

Rajendran, S., Nasir, S., 2014. ASTER spectral sensitivity of carbonate rocks – study in Sultanate of Oman. Adv. Space Res. 53, 656–673. Robson, D.A., 1971. The structure of the Gulf of Suez (clysmic) rift with special reference to the eastern side. J. Geol. Soci. London 127, 247–277. Rowan, L.C., Hook, S.J., Abrams, M.J., Mars, J.C., 2003. Mapping hydrothermally altered rocks at Cuprite, Nevada, using the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER). A new satellite-imaging system. Econ. Geol. 98 (5), 1019–1027. Rowan, L.C., Mars, J.C., 2003. Lithologic mapping in the Mountain Pass, California area using Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data. Remote Sens. Envir. 84 (3), 350–366. Sabins, F., 1997. Remote Sensing Principles and Interpretation, third ed., pp. 494. Salisbury, J.W., Walter, L.S., Vergo, N., D’Aria, D.M., 1991. Infrared (2.1-25 Micrometers) Spectra of Minerals. Johns Hopkins University Press, pp. 294. Shah, C.A., 2003. Independent component analysis mixture model (ICAMM) algorithm for unsupervised classification of multi/hyperspectral imagery. Master Thesis. Syracuse University, Syracuse, NY, pp. 1–188. Steckler, M.S., Berthelot, F., Lyberis, N., LePichon, X., 1988. Subsidence in the Gulf of Suez: implications for rifting and plate kinematics. Tectonophysics 153, 249– 270. Steen G., 1984. Radiometric age dating and tectonic significance of some Gulf of Suez igneous rocks. 6th Egyptian General Petroleum Corporation Conference, pp. 19. Sultan, M., Arvidson, R.E., 1986. Mapping of serpentinites in the Eastern Desert of Egypt by using Landsat Thematic Mapper data. J. Geol. 14, 995–999. Tyagi, P., Bhosle, U., 2011. Atmospheric correction of remotely sensed images in spatial and transform domain. Int. J. Image Proc. 5 (5), 564–579. Wang, Y.J., Lin, Q.J., Chen, Y., 2014. A method to improve mineral identification accuracy based on hyperspectral data. IOP Conf. Ser.: Earth Environ. Sci. 17. 012206. Xu, Y., Qizhong, L., Yun, S., Lu, W., 2004. Extraction mechanism of alteration zones using ASTER imagery. IEEE Trans. Geosci. Remote Sens., 4174–4175 Zoheir, B., Emam, A., 2012. Integrating geologic and satellite imagery data for high resolution mapping and gold exploration targets in the South Eastern Desert. Egypt. J. Afr. Earth Sci. 66–67, 22–34.

Please cite this article as: K. W. Abdelmalik, , The Egyptian Journal of Remote Sensing and Space Sciences, https://doi.org/10.1016/j.ejrs.2019.04.004