Space-Borne Imagery and Geochemical Characters of Post-Orogenic Dyke Swarms, Fatirah-Abu Zawal District, Eastern Desert of Egypt ()
1. Introduction
The Arabian Nubian Shield (ANS) is a compilation of Neoproterozoic juvenile arcs, younger sedimentary and volcanic basins, voluminous granite intrusions, and enclaves of pre-Neoproterozoic crust that crop out in the western Arabian Plate and the northeastern African Plate at the northern end of the East African Orogen [1] -[3] . Moreover, the final Pan-African phase of igneous activity in the Eastern Desert of Egypt, as a part of the ANS, is characterized by emplacement of bimodal Dokhan volcanics [4] , younger granitoids [5] , and numerous mafic to felsic dyke swarms [6] [7] . The dykes and dyke swarms are a common feature in the Neoproterozoic ANS terranes [8] . They represent a transitional period from compressional to extensional tectonic setting and their origin is controlled by lithospheric dynamics or passive rifting [6] [7] [9] [10] . In the Eastern Desert and Sinai Peninsula, these dykes are basaltic to rhyolitic in composition, closely associated with granitoids, and intruded under extensional conditions [7] [10] -[12] . The dyke swarms provide a window into the composition and evolution of the sub-continental mantle and hence the tectonic evolution of the overlying continental crust. Also, they may provide some geochronological controls on the later phases of crustal evolution and their orientation can provide information on the regional stress field at the time of emplacement [13] . Composite dykes that are made up of felsic and mafic rocks are abundant in the ANS, in particular in the Eastern Desert and Sinai of Egypt [14] -[16] . These dykes are incorporated in bimodal basalt-rhyolite dyke suites forming dyke warms up to tens of kilometers long. Regarding with El-Sayed [17] , three groups of dyke swarms have been distinguished in the Eastern Desert and Sinai: 1) metamorphosed syntectonic dykes (800 - 650 Ma; [18] ); 2) unmetamorphosed post-orogenic dykes (591 - 459 Ma; [15] [19] -[23] ) and 3) Neogene dykes (30 - 12 Ma; [24] [25] ) related to the updoming and opening of the Red Sea Rift system. The first two groups have variable composition ranging from mafic to felsic, whilst the Neogene dykes have the mafic composition. This paper contributes to characterizing the post-orogenic dyke swarms crosscutting the late Precambrian basement rocks at Wadi Fatirah-Wadi Abu Zawal District, Eastern Desert of Egypt (Figure 1). The current study has two essential aims. Firstly, it uses the remote sensing analysis of Enhanced Thematic Mapper (ETM+) data (174/42 path/row, acquisition date 09/03/2005) for detailed geological mapping and understanding tectonic setting of the dyke swarms. Band rationing and principal component transformation are the main remote sensing techniques applied to achieve this aim. Secondly, it aims to provide valuable petrological and geochemical data that help in understanding the origin of dyke swarms, their source component(s) and processes involved in magma genesis and evolution.
2. Geologic Setting
The exposed Precambrian rocks along Wadi Fatirah-Wadi Abu Zawal District (Figure 1) comprise metavolcanics (oldest), older granitoids, Dokhan volcanics, Hammamat sediments, younger gabbros and younger granites (youngest). Metavolcanics form a NE-SW elongated belt in the western part of study area. They are represented by island arc metavolcanics (=YMV of Stern, [26] ) that are characterized by bedding and lamination, and composed mainly of metabasalts, meta-andesites and their equivalent pyroclastics. Older granitoids occupy the eastern and southern parts of the mapped area and cover about 28% of total area. Generally, they are medium to coarse-grained rocks exhibiting low relief, spheroidal and cavernous weathering. Also, they locally possess foliation, particularly along their eastern margin. These rocks intruded through the metavolcanics and in turn they are intruded by the younger gabbros and younger granites. Therefore, older granitoids enclose xenoliths and roof pendants from the metavolvanics and are dissected by alkali feldspar pegmatite veins that are mined for feldspar ores, especially at Wadi El-Urf.
Dokhan volcanics occupy moderate to high relief terrains in the northwestern part of mapped area. They comprise andesites, quartz dacites, porphyritic rhyodacites and rhyolites where the latter two varieties increase in abundance westward. These volcanics are intruded by the younger granites and the felsic varieties are locally intercalated with the Hammamat sediments that occupy the northwestern corner of the mapped area. Hammamat sediments exhibit low to moderate relief and consist of polymictic conglomerates at the base grading upward into greywackes and siltstones. Younger gabbros crop out along Wadi Abu Zawal as homogeneous bodies that form terrains of low to moderate relief. They intruded the metavolcanics and older granitoids, and in turn they are intruded by the younger granites.
Younger granites comprise monzogranites, syenogranites and alkali feldspar granites that are arranged according to their relative age of emplacement. Monzogranites are buff in color and medium to coarse-grained and characterized by low to moderate relief, bouldery weathering and spheroidal exfoliation. They crop out in the
Figure 1. Geological map of Fatirah-Abu Zawal District, Eastern Desert of Egypt.
central part of the area and intruded into the older granitoids and metavolcanics. The contact against the older granitoids is gradational with a hybrid zone, up to 50m wide, whereas its contact with the metavolcanics is sharp and dip away from the granite intrusion. Syenogranites represent the dominant variety among the younger granites in the study area. They are mediumto coarse-grained and pale pink in color. Also, they form elongate bodies with moderate to high relief, intruded into the metavolcanics, older granites and younger gabbros, and consequently enclose xenoliths from the host rocks. Alkali feldspar granites have yellowish pink color and crop out in the northeastern part of the area under consideration. They form small plutons with oval to circular outlines and high relief (e.g. Gabal El-Dob, Figure 1).
The exposed Precambrian rocks in the area under consideration are crosscut by numerous post-orogenic dykes of mafic and felsic composition (Figure 2). The mafic dykes are the most abundant and widespread in the area comprising three petrographic varieties (basalt, basaltic andesite and andesite). On the other hand, the felsic dykes are petrographically distinguished into dacite and rhyolite. These dyke swarms are unmetamorphosed and arranged as closely parallel sets forming swarms, especially in the central and northern parts of the study area. They have length varying from few hundreds of meters up to 10 km and their width ranges from less than 1m up to 50 m. Also, the dykes are extending in dominant NE, NNE and ENE trends, while some dykes especially in southeastern part of the area exhibit NW trend. Contacts between the felsic and mafic dykes are commonly abrupt, without obvious signs of magma–magma interaction, but also without evidence of chilling. The bimodal dyke suite is the most abundant among the Neoproterozoic dykes in the Eastern Desert of Egypt. Younger granites, gabbros and older granites are the main host rocks of dyke swarms in the central, northern and eastern parts of the study area. Based on field observations and cross-cutting relations of the different dyke swarms, the dykes are straight indicating that they have been apparently emplaced along sharply defined fractures and they are emplaced contemporaneously.
Figure 2. Foliated calc-alkaline metavolcanics are intruded by (A) subparallel basalt to andeite and (B) acidic dyke swarms; (C) Acidic dykes intrude the granitoid rocks; (D) Acidic dyke is cut by basic one and both are intruded into the gabbroic rocks along Wadi Abu Zawal.
3. Petrography
Older Granitoids are Petrographically distinguished into quartz diorites, tonalites and granodiorites that grade imperceptibly into each other. Quartz diorites are medium to coarse-grained and characterized by hypidiomorphic granular texture, with weak foliation. These rocks are composed essentially of plagioclase, quartz, biotite, hornblende and very minor alkali feldspars. Sphene, zircon, apatite and opaques are the main accessories, whereas epidote, chlorite, sericite and kaolinite are secondary minerals. Tonalites are mineralogically and texturally similar to the quartz diorite, but differ in containing higher amounts of quartz. Granodiorites are coarse-grained and massive rocks, characterized by hypidiomorphic granular texture and composed mainly of plagioclase and quartz, together with alkali feldspars, biotite and hornblende. Zircon, apatite, sphene and opaque minerals are accessories, whilst chlorite, sericite and kaolinite are secondary minerals. In all varieties of older granitoids, Quartz fills the interstices between crystals of the early formed minerals, and displays wavy extinction as well as encloses zircon and apatite inclusions. Plagioclase possesses composition with anorthite content as (An20 - 35, An15 - 23, and An12 - 20) in the quartz diorites, tonalites and granodiorites, respectively. It occurs as euhedral to subhedral tabular crystals, commonly zoned with altered core and fresh rim. Its crystals are twinned according to the albite, pericline and albite-carlsbad laws. Orthoclase content exceeds 10% in the granodiorites. It forms subhedral to anhedral tabular and simply twinned crystals. Biotite forms subhedral flakes, occasionally altered to chlorite, strongly pleochroic, and enclose inclusions of quartz, apatite, zircon and sphene.
Younger granitoids in the area under consideration are microscopically differentiated into monzogranites, syenogranites and alkali-feldspar granites. Monzogranites are coarse-grained rocks, exhibit seriate porphyritic texture in the outer parts and equigranular hypidiomorphic texture in the core of intrusion. They are mainly composed of quartz, perthite, plagioclase, hornblende and biotite. Opaques, sphene, apatite and zircon are accessories, whereas sericite, chlorite and epidote are secondary minerals. Syenogranites are leucocratic in color, medium to coarse-grained and composed mainly of perthite, quartz, plagioclase, biotite and sometimes muscovite. Zircon, apatite and opaques are accessories, whereas secondary minerals include chlorite, sericite and epidote. Alkali feldspar granites are medium to coarse-grained, pinkish in color and essentially composed of orthoclase perthite, quartz, riebeckite, biotite and minor amounts of albite. Zircon, fluorite, cassiterite, apatite, allanite and opaques are accessories, whereas cholrite and sericite are secondary minerals. Alkali feldspars form subhedral tabular crystals and are represented by orthoclase perthites in both monzogranites and alkali feldspar granites. In syenogranites, alkali feldspars present as orthoclase and microcline perthites. Generally, perthites are of flame, string and patchy types. Plagioclase has composition with anorthite content as (An14 - 20, An12 - 15, and An5 - 10) in the monzo-syenoand alkali feldspar granites, respectively. Mica minerals in the studied younger granites are biotite and muscovite. Biotite occurs as anhedral to subhedral flakes that enclose fine zircon prisms. On the other hand, muscovite is recorded only in syenogranites, forming subhedral flakes and scaly aggregates. Hornblende is present only in the monzogranites and forms subhedral to euhedral prismatic crystals with apatite and zircon inclusions. Riebeckite is observed only in the alkali feldspar granites, as anhedral to subhedral crystals, strongly pleochroic from green to bluish green and enclose zircon, opaques and allanite inclusions.
3.1. Mafic Dykes
Basalt and basaltic andesite dykes are the most predominant rock varieties among the studied mafic dyke swarms. In hand-specimens, the mafic dykes are massive fine-grained, and display different color which varies from black to grayish green and greenish grey. Basalts and basaltic andesite dykes are composed mainly of plagioclase (labradorite composition (An55 - 65), augite and hornblende. Chlorite and actinolite are secondary minerals, while opaque oxides are accessories. They are characterized by ophitic to subophitic textures, besides porphyritic and intergranular textures. Plagioclase occurs either as tabular and prismatic phenocrysts (up to 3.8 ´ 2.7 mm) or as fine-grained subhedral laths (up to 1.4 mm long) in the groundmass. Euhedral to subhedral tabular plagioclase phenocrysts display lamellar twinning and and sometimes zoning with slightly altrered core and fresh rim. Pyroxene is represented mainly by augite which in turn occurs in the form of euhedral to subhedral prismatic phenocrysts and as granular aggregates in the groundmass. Augite is slightly altered to actinolite and chlorite along the crystal margins and cleavage planes.
Andesite dykes are less abundance compared to the basalt and the basaltic andesite dyke swarms. Megascopically, they are fine-grained, massive and display darkgrey to green color. It is commonly aphyric to faithfully porphyritic textures. In thin section, andesite dykes are composed mainly of plagioclase, and few hornblende phenocrysts, embedded in fine grained groundmass consists of plagioclase laths, chlorite, opaque oxides, calcite and saussurite. Porphyritic and amygdaloidal textures are present, where the amygdales are frequently filled with calcite and/or cryptocrystalline silica. Plagioclase (An25 - 45) is slightly altered to sericite and/or epidote. It occurs as subhedral to anhedral phenocrysts up to 2.5 ´ 3.8 mm, and as fine-grained subhedral laths that constituting the groundmass. In some slices, plagioclase displays zoning with altered core and fresh rim. Quartz, if present, occurs as anhedral fine-grained crystals in the groundmass.
3.2. Felsic Dykes
Dacite dykes are buff to reddish brown color, fine-grained, massive and exhibit porphyritic and less commonly aphyric textures. They are composed mainly of quartz and plagioclase phenocrysts, embedded in fine grained groundmass formed of plagioclase laths, quartz, and opaques. Accessory minerals are opaques, apatite, and zircon, whereas epidote and chlorite are secondary minerals. Plagioclase (An10 - 18) occurs as euhedral to subhedral prisms (up to 2 ´ 3.7 mm) and partially altered to epidote. Euhedral to subhedral K-feldspar phenocrysts (up to 2.5 ´ 3.9 mm) and slightly altered to kaolinite ans sometimes sericite. Quartz occurs as large subhedral phenocrysts and fine-grained crystals filling the interstices between the other constituents.
Rhyolite dykes are pink to buff color, fine-grained and massive. Microscopically, the rhyolite is composed mainly of quartz and orthoclase phenocrysts, plagioclase (An < 10), embedded in cryptocrystalline and glassy groundmass that speckled with quartz, feldspars and opaques. Zircon, and opaques are the accessory minerals, while epidote, chlorite and sericite are the alteration products. K-feldspar is represented by euhedral to subhedral tabular and prismatic phenocrysts as well as small laths in the groundmass. Quartz occurs as subhedral phenocrysts (up to 2.5 mm across) and it is frequently fills the interstices between other minerals. Plagioclase forms Euhedral to subhedral prismatic phenocrysts (up to 2.6 ´ 3.8 mm) and exhibit lamellar twinning and sometimes zoning. Plagioclase is slightly altered to sericite and epidote. Generally, rhyolite dykes are characterized by porphyritic, glomeroporphyritic and spherulitic textures.
4. Remote Sensing Data Analysis and Interpretation
Remotely sensed data are important source of information needed to achieve purposes of the geological studies, especially geological mapping, mineral exploration and structural interpretation. Several digital image processing techniques (image enhancement, fusion, rationing, and principal component transformation) were employed for lithological mapping and identifying mineral deposits [27] -[42] . However, there are two important factors that should be taken in consideration during lithological mappings by using remote sensing techniques. First is the localization of increased concentration of minerals relative to the background, and second is the characterization of the mineral assemblages [43] . The present study uses subset of corrected Landsat ETM+ data to discriminate between the different lithologic units, the different dyke varieties and delineate the structural lineaments pattern at the Wadi Fatirah–Wadi Abu Zawal area. Remote sensing techniques including band composites, band ratios and principal component transformation are employed in the present research.
The most useful RGB false color combination (FCC) of ETM+ bands is 7-4-2 which exhibits regional structural lineaments, joints and faults [44] . Consequently, the false color composite band image (7-4-2) was created for the study area (Figure 3), where it powerfully separates the dyke swarms from their host rocks. Moreover, this FCC image enhances the drainage pattern and structural lineaments features.
Principal component analysis (PCA) is a multivariate statistical technique that selects uncorrelated linear combinations (eigenvector loadings) of variables in such a way that each successively extracted linear combina-
Figure 3. False color composite band image (7-4-2) in RGB mode, enhances the drainage lines, structural lineaments and dyke swarms in Fatirah-Abu Zawal area.
tion, or principal component (PC), has a smaller variance [45] . On the other hand, band rationing of Landsat ETM+ data enhances the spectral differences between bands and reduces the environmental and topographic effects. The band ratio images are prepared simply by dividing the DN values of each pixel in one band by the DN values of another band [46] . Therefore, band ratios have been successfully used for lithological mapping and discriminating between different rock types in the Eastern Desert of Egypt (e.g. [47] -[49] ). Also, several studies around the world have used the PCA and band rationing techniques of TM data for mapping hydrothermal alteration zones and lithologies (e.g. [27] [32] [36] [40] [50] -[61] ).
The PCA transformation is carried out for the six ETM+ bands (1, 2, 3, 4, 5 and 7) of the study area. Visual examination of the generated FCC PCA images in this study revealed that the RGB combinations (PC3, PC1, PC4) and (PC4, PC2, PC3) are the most informative images. On the hand, (5/7, 4/5, 3/1) and (5/7, 5/1, 5/4 × 3/4) RGB ratio images are the most valuable combinations. Hence, they are used in the present study for discriminating the different varieties of dyke swarms and mapping the different rock units in the study area.
On FCC PCA (PC3:R, PC1:G, PC4:B) image, the dyke swarms can be easily discriminated (Figure 4). The mafic dykes have purple image signature whereas the felsic dykes have blue and light blue tone. Dokhan volcanics and alkali-feldspars granites have dark blue and light bluish image signatures, respectively. On contrast, gabbros and basic metavolcanics exhibit pinkish and light purple colors. This FCC is failed to discriminate between monzogranites and syenogranites, where both have greenish color. Also, both gabbros and metavolcanics exhibit the same image signature. Consequently, FCC PCA (PC4: R, PC2:G, PC3:B) image was processed (Figure 5) to clearly differentiated between mozogranites (greenish) and syenogranites (dark blue) as well as between gabbros (pinkish) and metavolcanics (light blue).
FCC ratio image (5/7: R, 4/5:G, 3/1:B), discriminates easily between the dyke swarms (Figure 6), where the mafic dykes have reddish image signature and the felsic dykes show green and bluish green tone. Dokhan vol-
Figure 4. FCC PCA image (PC3, PC1, PC4).
Figure 5. FCC PCA image (PC4, PC2, PC3).
Figure 6. FCC ratio image (5/7, 4/5, 3/1).
canics show lemon color, whereas alkali-feldspars granites have pinkish image signature. On the other hand, gabbros and basic metavolcanics exhibit red and light green colors, respectively. Both monzogranites and syenogranites have light blue image signature. On the FCC (5/7:R, 5/1:G, 5/4 × 3/4:B) image, the mafic dykes have purple image signature whereas the felsic dykes have bluish color (Figure 7). Dokhan volcanics and alkalifeldspars granites have reddish and yellowish image signatures, respectively. Gabbros and basic metavolcanics exhibit pinkish and purple colors. Monzogranites and syenogranites have green and light greenish colors, respectively. The FCC (4/2:R, 5/7:G, PC3:B) image was processed to clearly differentiated between the dyke swarms and different rock units occupying the study area (Figure 8). The mafic dykes show light bluish color,
Figure 7. FCC ratio image (5/7, 5/1, 5/4 × 3/4).
Figure 8. FCC image (4/2, 5/7, PC3) in red, green and blue channels, respectively.
whereas the felsic dykes exhibit orange and reddish colors. Dokhan volcanics have light green color, alkali-feldspars granites have yellowish image signature, while gabbros and basic metavolcanics exhibit light blue and bluish colors. Monzogranites and syenogranites have reddish and orange image signatures, respectively.
5. Geochemistry
5.1. Analytical Techniques
The present study is based on 45 samples collected mainly from the different varieties of dyke swarms in Fatirah-Abu Zawal district, Eastern Desert, Egypt. During the detailed microscopic study for the collected samples, the least altered and weathered samples were selected for major and trace elements analyses. The major and trace elements were determined by using X-ray Fluorescence Spectrometer (XRF) within the Central Laboratories of the Geological Survey of Egypt.
5.2. Major and Trace Element Characteristics
Thirteen representative samples were analyzed (4 samples from basic dykes, 4 samples from andesitic dykes, and 5 samples from acidic dykes) for major and some trace elements. The chemical compositions of representative samples and their CIPW norms are listed in (Table 1), moreover some compositional ratios and indices were calculated and listed in (Table 2). In the acidic dykes, the SiO2 content ranges from 73.92% to 77.35%, that of total alkalis (Na2O + K2O) from 7.48% to 8.97%, CaO from 0.40% to 0.62%, Na2O/K2O ratio from 0.63 to 1.23 and MgO/(FeOt + MgO) ratio from 0.06 to 0.15. The andesitic dykes show SiO2 content ranging from 54.1% to 60.62%, (Na2O + K2O) from 3.14% to 7.26%, CaO from 3.41% to 7.3%, Na2O/K2O ratio from 0.62 to 1.57 and MgO/(FeOt + MgO) ratio from 0.27 to 0.46. On the other hand, the basic dykes have SiO2 content ranging from 48.59% to 52.89%, (Na2O + K2O) from 3.41% to 3.92%, CaO from 5.11% to 7.73%, Na2O/K2O ratio from 2.22 to 3.39 and MgO/(FeOt + MgO) ratio from 0.27 to 0.49. It is obvious that the contents of SiO2, Na2O, K2O, Ba, Y, Rb and Th are increased from basic to acidic dykes, and vice versa relative to the contents of CaO, MgO, Fe2O3, Sr, V, Co and Ni. Thoronton and Tuttle [62] defined the differentiation index (D.I) as the sum of weight percentage of normative quartz, orthoclase, albite, nepheline, leucite and kasilite. Nocklods (1954) defined the mafic index (M) as (FeO + Fe2O3)*100/( FeO + Fe2O3 + MgO), while the Mg# index equals [(mol. Mg × 100)/(Mg + Fe)], where each group of rocks are crystallized from magma with distinct Mg# values (De Bari, 1994). The mafic dykes show Mg# values ranging from 42.1 to 65.27, while the acidic dykes have smaller Mg# values (10.54 - 26.1).
The relationship between the different varieties of dyke swarms can be investigated from the variation diagrams of some major and trace elements against the SiO2 (Figure 9 and Figure 10). The regular variation among the plotted major and trace elements of the basic dyke swarms indicate that the fractional crystallization processes may play significant role in their formation. The CaO, MgO, Fe2O3, Sr, V, Co and Ni contents are decreased with increasing SiO2, while SiO2, Na2O, K2O, Ba, Y, Rb, Zr and Th are increased with increasing SiO2. This feature is consistent with the crystallization of pyroxene, plagioclase, magnetite and amphibole in the early formed basaltic dykes. Moreover, the increase in the Na2O, K2O and SiO2 reflect the increasing of acidity throughout the acidic dykes. Also, these dyke swarms have mostly low Mg# value indicating that they were produced from highly evolved magma. The MgO/(MgO+FeO) molar ratios (Mg#) for the mafic dykes decrease gradually from basalt through andesite to the most evolved rhyolite. The Na2O content is generally higher than K2O in the mafic and felsic dykes.
On the variation diagrams, it is observed that there is a conspicuous compositional SiO2 gap exists between mafic dykes (basalt and andesite dykes) on the one hand and the felsic dykes on the other hand. Moreover, there is a distinct two trends for the studied dyke swarms one for the mafic dykes and the second for the felsic dykes. Therefore, the presence of a compositional gap and the two trends between them may suggest independent origin (i.e. are not comagmatic). Moreover, on these diagrams, the felsic dykes, characterized by low abundance of incompatible elements (Table 1), and plots as a separate cluster in almost all covariation diagrams (Figure 9 and Figure 10) and confirmed the observation that they appear not to be related to the mafic dykes via fractionation and suggest that each suite requires different primary melts. Incompatible trace-element variation patterns for the mafic and felsic dykes, normalized to the N-Type MORB [63] , are shown in (Figure 11). The mafic dykes show an inclined smooth pattern (i.e. enrichment in LILE relative to HFSE) and are characterized by
Table 1. Major oxides (%), trace elements (ppm) and CIPW norms of dyke swarms.
Continued
Table 2. Compositional ratios and indices calculated for dyke swarms.
negative Zr and Ti anomalies. The felsic dyke samples illustrate similar pattern to the mafic dykes but with higher HFSE contents.
5.3. Chemical Classification and Magma Type
The (Na2O + K2O) versus SiO2 diagram [64] is used to approach the nomenclature of the studied dyke swarms (Figure 12(a)). It is obvious that the plotted samples of basic dykes fall within basalt and basaltic andesite fields, those of intermediate dykes fall within basaltic andesite and andesite fields and one sample falls within the trachyandesite field, while those of acidic dykes are plotted within rhyolite field, giving harmony with the petrographic description. This is again supported by plotting the representative samples on the SiO2 versus Zr/TiO2 × 0.001 diagram [65] , where the plots of basic, intermediate and acidic dykes fall within sub-alkaline basalt, andesite and rhyolite fields, respectively (Figure 12(b)). It is observed that the studied mafic and felsic-
Figure 9. Variation plots of K2O, CaO, Fe2O3, MgO, Al2O3 and Na2O versus SiO2. Symbols as: (¯) Basaltic dykes, (¿) andesitic dykes, (q) acidic dykes.
dyke swarms display a wide compositional range: 1) a mafic dyke swarms (48 - 60 wt% SiO2) which include basalt, basaltic andesite and andesite and 2) a felsic dyke swarms (73 - 77 wt% SiO2) that comprising dacite, and rhyolite.
The analyzed samples of dyke swarms have been plotted on the Na2O+K2O versus SiO2 binary diagram [66] , where all plots of studied dykes fall within the sub-alkaline field (Figure 12(c)). This sub-alkaline nature of
Figure 10. Variation plots of Rb, Sr, V, Ba, Ni, Co, Cr, Nb, Y and Zr versus SiO2. Symbols as in Figure 9.
these dyke swarms may be supported by increase in SiO2, Na2O and K2O, and decrease in MgO/(FeOt+MgO) ratio from the basic to acidic dykes. On the AFM diagram [67] , a well-defined tholeiitic to calc-alkaline trend is observed (Figure 12(d)). The basic dykes plot away from the MgO apex within tholeiitic field, whereas the andesitic dyke plots mostly fall within calc-alkaline field, except one sample falls within tholeiitic field. On the other hand, all acidic dyke plots fall within calc-alkaline field along the FeOt-total alkalis join in a cluster close to the total alkali apex. This reflects the very low MgO and high alkalis contents in the acidic variety of dyke swarms. Miyashiro [68] suggested that relationship between FeO*/MgO as fractionation index and TiO2 is diagnostic of tholeiitic and calc-alkaline suites. Therefore, plotting of the analytical data on the TiO2 versus FeOt/MgO diagram [68] shows that the basic and andesitic dykes follow the tholeiitic trend (Figure 13(a)), meanwhile the acidic dykes extends to the calc-alkaline trend, out of diagram limits due to their high FeOt/MgO values. From the above discussion, it is clear that the basic dykes have tholeiitic affinity and the andesitic dykes
Figure 11. Spider diagram showing the average incompatible elements of studied dyke swarms, normalized to N-Type MORB [63] . Symbols as in Figure 9.
Figure 12. Plots of studied dyke swarms on (A) TAS diagram [64] ; (B) Zr/TiO2-SiO2 [65] ; (C) Na2O+K2O vs. SiO2 [66] and (D) AFM ternary diagram [67] . Symbols as Figure 9.
show tholeiitic to calc-alkaline affinity, while the acidic dykes exhibit calc-alkaline characters.
5.4. Tectonic Setting
The analytical data of dyke swarms are plotted on the Y vs Nb and Y + Nb vs Rb diagrams (Figure 13(b) and Figure 13(c)) after Pearce et al. [69] . The basic and intermediate dykes fall mostly within the field of volcanic arc environment, whereas the acidic dykes display geochemical features of within plate rocks. On the Th-Nb/16- Hf/3 ternary diagram [70] , all plots of dyke swarms fall within the field of destructive plate margin basalts and differentiates (Figure 13(d)).
6. Discussion and Conclusion
The Wadi Fatirah-Wadi Abu Zawal District represents the northern most edge of the ANS that formed during the Neoproterozoic (ca.900 - 550 Ma) as part of the East African Orogen by accretion of juvenile arc terrains. The Late Neoproterozoic crustal evolution (ca.610 - 540 Ma) in the ANS occurred in an extensional tectonic environment [7] [71] [72] and was accompanied by abundant igneous activity: emplacement of alkaline granitoids, eruption of bimodal volcanics, and intrusion of bimodal dyke swarms [1] [6] [73] -[75] . In such extensional tectonic environments, ascent of mantle-derived mafic magma into the crust could either melt the existing continental crust and mix with the partial melts or merely result in fractional crystallization of mafic magma leading to granite formation [76] . The plotted samples on the variation diagrams for major and trace elements (Figure 9 and Figure 10) suggest that there are significant geochemical differences between the mafic and felsic dyke swarms. The most effective geochemical differences are the highest content of the mafic dyke swarms in the
Figure 13. Plots of studied dyke swarms on (A) TiO2 vs. FeOt/MgO diagram [68] ; (B) Y − Nb diagram [69] ; (C) Y + Nb vs Rb diagram [69] and (D) Th-Nb/16-Hf/3 ternary diagram [70] . Symbols as Figure 9.
HFS elements relative to the felsic dykes. The chemical differences suggest that the mafic and felsic dykes cannot be related and there is no chemical evidence for their generation from the same magma source, as previously argued by Stern and Gottfried [73] for the Eastern Desert dykes. Moreover, the wide variations in the composition of the mafic dykes are attributed to chemical and mineralogical heterogeneities in the mantle sources, differential partial melting of the source materials and modifications of magma composition by fractional crystallization and/or crustal contamination processes. The ratios of Zr/Ti, Zr/Y and Zr/ Nb are not change significantly during the generation of the mafic magma and therefore are very important for evaluation of the relative roles of fractional crystallization, partial melting, and source heterogeneity in producing the variations in magma composition [77] [78] . The presence of negative Ti anomalies for the mafic dykes (Figure 11) can have several causes such as amphibole and/or Ti-magnetite fractionation and selective transport in fluids or prior extraction of melts.
7. The Following Are the Main Conclusions
The dyke swarms of Wadi Fatirah-Wadi Abu Zawal District are differentiated into two petrogenetic groups; namely as mafic and felsic dykes. The mafic dykes comprise basalt, basaltic andesite and andesite, while the felsic dykes are dacite and rhyolite in composition.
Band rationing and PCA are two effective remote sensing techniques applied on ETM+ data for detailed lithological mapping and spectral discrimination of the dyke swarms of Wadi Fatirah-Wadi Abu Zawal District.
The RGB (PC3, PC1, PC4) and (PC4, PC2, PC3) PCA images and RGB (5/7, 4/5, 3/1) and (5/7, 5/1, 5/4 × 3/4) ratio images are the most informative images used to clearly discriminate the different varieties of dyke swarms as well as other lithologies in the study area.
Geochemically, the mafic dyke swarms show tholeiitic to calc-alkaline affinity, with a negative Ti anomaly in their normalized plots of incompatible elements. In contrast, the felsic dyke swarms display significant calc-alkaline affinity. They show decreasing LILE contents and increasing amounts of Y, Nb, and Zr.
The dyke swarms have been emplaced in destructive plate margin settings during periods of extension. The basic dyke swarms show characteristics of volcanic arc setting, while the acidic dykes display geochemical features of within plate rocks.
The chemical differences between the mafic and felsic dyke swarms favor that these two types of dyke swarms cannot be related to the same magma source, through fractional crystallization processes, but they have been formed from two different parental magmas.