sign in sign up
<<
Home , Gypsum

S Journal of

O p s e s n Acce Geology and Geoscience

Research Article Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern , Using ASTER Thermal Data

Habes A. Ghrefat1* 1Department of Geology and Geophysics, College of Science, King Saud University, Riyadh, Saudi Arabia

Abstract Gypsum is of great commercial value and is used in different agricultural and industrial applications. The GER 3700 spectroradiometer (0.3-2.5µm) was used to measure the laboratory spectra of the samples. Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) (AST_05) data (8-12 µm) were used to detect and map the gypsum deposits in the Midyan area of northwestern Saudi Arabia. These data were processed using the minimum noise fraction, pixel purity index, and nD visualization to derive image end members. Gypsum spectrum showed diagnostic absorption features located at 1.4, 1.9, 2.2, and 8.63µm. The spatial maps of gypsum deposits were determined using band ratio (BR) (band 12/11) and matched filtering (MF) techniques. The results of BR and MF match well with the published geologic map and field data. The results of this study is of economic importance in hyper-arid environments such as Saudi Arabia. Keywords: ASTER, Thermal, Saudi Arabia, Gypsum, mapping

Introduction mapped using decorrelation stretching (DECOR) applied to Landsat 5 Thematic Mapper (TM) data. Airborne Visible/ are deposits or unconsolidated salts Infrared Imaging Spectrometer (AVIRIS) has been used to precipitated out of supersaturated water solutions. Precipitation identify hydroboracite, pinnoite, rivadavite, and damp is induced by removing water and concentrating the solution in Death Valley, California using the least squares spectral by evaporation or freeze-drying [1]. According to and, feature fitting method. used the reflectance spectra of visible evaporites represent a distinct and fairly abundant group of and near-infrared (VNIR) through shortwave infrared (SWIR) economic deposits of sedimentary origin [2, 3]. These deposits (0.4-2.5 µm) light to study 35 saline in playa are closely interrelated in their environmental and geological minerals. Landsat 5 Thematic Mapper was used to conditions of sequential formation, geographic occurrence, map the spatial distribution of gypsum and halite in the Chott physical and chemical properties, and industrial use. el Djerid salt playa, Tunisia. The proportional abundance Gypsum (CaSO4.H2O) is of great commercial value, and is maps of these end members were determined using the linear one of the most useful and widely used minerals of the 20th mixture model. used the Thermal Infrared Multispectral century [4]. Large quantities of gypsum are consumed used Scanner to map playa evaporite minerals in Death Valley, in the construction , such as in , plaster of California, and successfully differentiated gypsum, halite, and , wallboard, , and ceramic tiles. Gypsum is used thenardite . Gypsum minerals were mapped in using as an amendment to neutralize sodic in the agricultural the band rationing, DECOR, principal component analysis, activities. Most gypsum in the is used to make feature-oriented principal components (FOPCA) methods, wallboard for homes, offices, and commercial buildings. The and sulphate and thermal indices applied to ASTER data [12]. four varieties of gypsum include , satin , ; The results showed that the FOPCA and DECOR methods and roses. provided better results than the other methods (Esra, 2019). Soltaninejad et al. (2018) applied the spectral angle mapper Measurements of the surface of the Earth using thermal (SAM) method to ASTER and Landsat 8 to map evaporite infrared (TIR) are important for identification of the chemical minerals in Sirjan playa, Kerman. composition of rocks and soils. Silicate and non-silicate minerals display diagnostic absorption features in the TIR region [5, 6]. Evaporite minerals were detected and mapped Correspondence to: Habes Ghrefat, Department of Geology and Geophysics, using different types of remote sensing data [7, 8, 9-11]. The King Saud University; Email: habes[AT]ksu[DOT]edu[DOT]sa evaporite mineral zonation in salt flats in Chile has been Received: Dec 18, 2019; Accepted: Jan 17, 2020; Published: Jan 20, 2020

J Geol Geosci 1 Volume 4(1): 2020 Ghrefat HA (2020) Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern Saudi Arabia, Using ASTER Thermal Data

To the best of the author’s knowledge, this is the first study is part of the Maqna massif. It is bounded to the west by steep to use reflectance spectroscopy and thermal remote sensing cliffs along the Gulf of Aqaba and by the Wadi Ifal plain to the for mapping gypsum deposits in the Midyan area. This study east and south. The massif has a steep relief of several hundred aims to use AST_05 data (8-12 µm) to detect and map the meters and is incised by numerous wadis. The highest point is distribution of evaporite deposits based on their emittance Jabal Hamdah at 710 m above sea level. spectra. Different techniques, including atmospheric correction, minimum noise fraction (MNF), nD visualization, The deposit is part of the thick evaporite, carbonate, and pixel purity index (PPI), band ratio, and matched filtering clastic sedimentary Miocene Raghama Formation. The group (MF), were applied to the ASTER thermal data. The results occupies the flanks of a northeasterly oriented, 45 km long, of this study will be of economic importance in hyper-arid south-plunging pericline that has a core of Proterozoic granitic environments such as Saudi Arabia. rocks. The evaporites are found mainly in the bed formation (Middle to Late Miocene) in the uppermost part of the Materials and methods Raghama Formation. The formation varies in thickness from Area of study: The Maqna gypsum deposits have an outcrop approximately 150 m to 300 m. The changes in thickness reflect area of approximately 270 km2 in the Midyan area of the paleogeography of a continent-influenced sedimentary northwestern Saudi Arabia. The principal deposit is on Jabal environment to the north and marine deposition to the south ar Raghamah (Rughamah) (28°24′N, 34°54′E) approximately and east. The overall evaporite thicknesses are greatest in the 160 km west of Tabuk. Another important occurrence is inland south and east, and the beds are progressively attenuated and from Maqna on the Gulf of Aqaba coast (Figure 1). The area eventually disappear northward. The formation is overlain

Figure 1: Location map and geologic setting of the study area.

J Geol Geosci 2 Volume 4(1): 2020 Ghrefat HA (2020) Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern Saudi Arabia, Using ASTER Thermal Data

by poorly consolidated deposits of the clastic Pliocene Lisan spatial resolution, the identification of a variety of surface Formation. materials is possible. Open fold structures in the bed formation might have resulted Image processing of ASTER thermal data: AST_05 data (bands from diapiric movement of the gypsum/, and slump 10-14) were used in the current study. These bands were first structures are evident. The commonly steep four major cycles transformed using the MNF transformation in order to estimate are separated by interludes of marl and dolomite deposition. and remove noise, determine the information dimensionality of The thickest individual sulphate bed is 110 m thick, the imagery, and facilitate the subsequent computer processing and the cumulative thickness of the successive beds is and analyses [15]. The PPI was run on the MNF images of these approximately 140 m. datasets to isolate the purest or least-mixed images by projecting Gypsum exposed at the surface gives way to anhydrite at MNF-transformed image spectra to random unit vectors until approximately 15 m to 20 m depth; this suggests that the the most extreme pixels were counted and tallied [16]. The nD gypsum was formed by rehydration of anhydrite by meteoric visualizer was then used to cluster isolated spectra into discrete waters. The analysis of cores from shallow drill holes indicated PPI image endmember classes [17]. that anhydrite/gypsum ratios can vary even in the upper 15 m Mapping methods: to 20 m with hydration of anhydrite to gypsum ranging from partial to almost complete, but seldom exceeding 90%. It is Matched filtering: MF is a classification method that probable that variations in the original crystallinity of the maximizes the response of known endmembers and suppresses anhydrite, possibly due to depositional factors, controlled the the response of the composite unknown background using least change to gypsum. Both gypsum and anhydrite vary in texture square regression methods. It is based partly on foreground- from finely aphanitic to coarsely crystalline, and contain few background spectral unmixing [18]. The identification of impurities. materials is based on matches to library or image endmember spectra. MF can best be described as partial unmixing because Characteristics of remote sensing data: ASTER was launched it is not constrained by the unit sum rules of linear spectral on the platform of the Terra spacecraft of NASA in 1999 unmixing. The results of MF provide a means of estimating [13, 14]. ASTER is a multispectral sensor with high spatial the relative degree of matching between a reference and image resolution (15 m VNIR) and extensive spectral coverage from spectra where ≥1 is a perfect fit. the visible to TIR region (Table 1). ASTER is comprised of three optical subsystems, namely VNIR, SWIR, and TIR Band ratioing: BR is used to effectively display spectral radiometers. ASTER has three bands in the VNIR, six bands variations [19, 20]. BR is used to enhance the contrast between in the SWIR, and five bands in the TIR with 15, 30, and 60 materials by dividing the brightness values at peaks/maxima m spatial resolutions, respectively. Because ASTER data are and troughs/minima in a reflectance curve, after removing characterized by wide spectral coverage and relatively high the atmospheric effects. Spectral band rationing enhances

Table 1: ASTER data properties. Sensor Subsystem Band No. Spectral range (µm) Quantization levels (bits) Spatial resolution (m) Swath width (km) 1 0.52-0.60 VNIR 2 0.63-0.69 3N 0.78-0.86 15 15 3B 0.78-0.86 ASTER 4 1.600-1.700 60 5 2.145-2.185 30 SWIR 6 2.185-2.225 7 2.235-2.285 8 2.295-2.365 9 2.360-2.430 8 10 8.125-8.475 12 11 8.475-8.825 TIR 12 8.925-9.275 90 13 10.25-10.95 14 10.95-11.65

Country 2016 2017 2018 United States 17,000 20,700 21,000 130,000 15,500 16,000 16,000 16,000 16,000 Turkey 9,000 16,100 21,000 Saudi Arabia 1,860 3,150 3,200

J Geol Geosci 3 Volume 4(1): 2020 Ghrefat HA (2020) Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern Saudi Arabia, Using ASTER Thermal Data

compositional information while suppressing other types of Table 2: XRF results of the collected samples. information about earth’s surface, such as terrain slope and Sample Wt. % Gypsum Granite grain size differences [21]. SiO2 46.7 73.0 64.4

Laboratory spectral measurements: The collected samples Al2O3 12.1 3.7 15.3 Fe O 5.9 4.4 3.8 were spectrally measured in the laboratory using a GER 3700 2 3 spectroradiometer (Geophysical and Environmental Research CaO 1.3 3.9 3.8 Corporation, 1999). An external light source (a 500 W halogen MgO 11.5 3.1 0.6 Na O 1.0 0.7 5.6 lamp) was used for performing the spectral measurements. The 2 K O 2.5 1.1 2.9 instrument is made up of 640 spectral bands between 0.315 2 Cr O 0.0 0.0 0.0 µm and 2.519 µm, and the field of view of the instrument is 2 3 TiO 0.8 0.6 0.6 approximately 100 cm2. 2 MnO 0.1 0.2 0.1

Results and discussion P2O5 0.6 0.3 0.1 SrO 0.0 0.0 0.0 The United States Geological Survey (USGS) Library and BaO 0.1 0.0 0.1 GER 3700 spectroradiometer spectra of gypsum, sandstone, LOI 17.5 9.1 2.7 and granite is shown in Figure 2. Major absorption features in the VNIR and SWIR regions distinguished the spectrum of gypsum [22, 23]. The absorption feature around 1.2 µm is due to a combination of the H-O-H bending and the first overtone of O-H stretching. The absorption features between 1.4 µm and 1.6 µm are due to the first overtone of O-H stretching. The absorption bands near 1.74 µm are due to combinations involving H-O-H bending, O-H stretching, and low-frequency vibration modes of the structural water molecules. The strong absorption features near 1.9 µm are due to a combination of O-H stretching and H-O-H bending. The absorption bands around 2.2 µm are attributed to a combination of O-H stretching and the first overtone of water. The granite spectrum was characterized by several absorption features in the VNIR-SWIR range (Figure 2). The presence of iron in granite samples cause absorption features located between 0.40 µm and 0.60 µm and at 0.87 µm. XRF results showed that Fe203 in the collected samples range from 1.4 to 10.4% (Table 2). The absorption feature located at 2.20 µm is due to the presence of Al-OH, whereas that at 2.35 µm was due to Fe, Mg-OH. The percentages of Al203 and MgO in granite samples range from 12-19%, and 0.3 -0.9 %, respectively. The sandstone spectrum displayed absorption features at 0.5, 1.4, Emittance spectra of gypsum, granite, and 1.9, and 2.2 µm owing to the presence of Fe, water, and Al- Figure 3: sandstone. OH, respectively [24-26]. The emittance spectra extracted from AST_05 data is shown in Figure 3. The spectrum of gypsum is characterized by a low emissivity value located at ASTER band 11 (8.63 μm). The other bands have higher emissivity value. The spectrum of granite is distinguished by deep and strong absorption feature located at ASTER band 12 (9.1 μm). This feature is helpful for discriminating other felsic rocks such as granodiorite and rhyolite. The spectrum with absorption features at ASTER bands 10 (8.3 μm) and 12 (9.1 μm) was similar to the sandstone spectrum. The results showed a good match between the spectra derived from AST_05 data and those obtained from the converted Johns Hopkins Library (JHU) emissivity spectral library. Figure 2: The laboratory measured spectral reflectance of gypsum, granite, and sandstone using the GER 3700 Figure 4 shows the color composite image of AST_05 data spectroradiometer. (R=band 11, G=band 12, B=band 10) of the study area. In

J Geol Geosci 4 Volume 4(1): 2020 Ghrefat HA (2020) Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern Saudi Arabia, Using ASTER Thermal Data

Figure 4: ASTER 05 image of the study area (R=12, G=10, B=11).

Figure 6: ASTER band ratio (12/11) of gypsum deposits.

The abundance estimates for gypsum deposits using MF method is shown in Figure 5. Gypsum fractions range from 0.23 to 1.0. Brighter pixels in the abundance images represent end members with higher abundances. The MF results showed that the gypsum fraction was higher in the northeastern and southwestern parts of the study area. The MF and BR (Figure 6) results revealed similar patterns of distribution of gypsum deposits. They were useful for distinguishing and mapping gypsum deposits based on their emittance spectra. The data and methods used in this study are of great importance for mineral exploration in arid and semi-arid areas. Conclusion The TIR bands of AST_05 data were useful for detecting and mapping the gypsum deposits in the study area. Image-derived spectra of the gypsum deposits showed an absorption feature located at 8.63 µm. The band ratio and MF methods obtained a similar spatial distribution pattern of the gypsum deposits. The fractional abundance of gypsum deposits was greater in the southeastern and northeastern parts of the study area. Good correspondence between BR and MF and the geological Figure 5: Abundance (fraction) image of gypsum deposits using MF. map of the study area, and the field details observed. Partial or unconstrained unmixing methods, such as MT, are more robust for use with image end members. The results showed that there this image, gypsum deposits are mainly displayed in the light was a good match between the representative image spectra blue pixels. The gypsum deposits were mainly distributed of gypsum derived from ASTER and the corresponding JHU in the northeastern and southwestern parts of the study area. emissivity spectra library for gypsum. The thermal bands of The whitest zones in the 12/10 ratio image represents areas ASTER data are useful for mineral exploration studies in arid dominated by gypsum deposits, while granitic rocks and clastic and semi-arid regions such as Saudi Arabia. sedimentary rocks are showed in the dark pixels. Gypsum has References higher emissivity value (band 12) and lower emissivity value 1. Sonnenfeld P, Perthuisot JP (1989) and Evaporites. (band 10) (Figure 3). American Geophysical Union. [View Article]

J Geol Geosci 5 Volume 4(1): 2020 Ghrefat HA (2020) Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern Saudi Arabia, Using ASTER Thermal Data

2. Reeves CC (1978) Economic significance of playa lake deposits, 15. Hunt GR, Salisbury, JW, Lenhoff CJ (1973) Visible and near- Separation Public. Int Assoc Sedimentol 2: 279-290. [View Article] infrared spectra of minerals and rocks. IV. and . Mod Geol 4: 85-106. [View Article] 3. Sonnenfeld P (1984) Brines and Evaporites. Academic Press. [View Article] 16. Boardman JW, Kruse FA, Green RO (1995) Mapping target signatures via partial unmixing of AVIRIS data. JPL Publication 4. Kyle JK (1992) Evaporites Evaporitic Processes and Mineral 1: 23-26. [View Article] Resources. In: Melvin, J.K (Ed.), Evaporites, Petroleum and Mineral Resources. Elsevier. [View Article] 17. Boardman JW (1993) Automated spectral unmixing of AVIRIS data using convex geometry concepts. JPL Publication 1: 11-14. 5. Gupta RP (2003) Remote Sensing Geology. Edn Springer. [View [View Article] Article] 18. Smith MO, Johnson, PE, Adams JB (1985) Quantitative 6. Hook SJ, Abbott EA, Grove C, Khale AB, Palluconi FD (1999) determination of mineral types and abundances from reflectance Use of Multispectral Thermal Infrared Data in Geological Studies. spectra using principal components analysis. J Geophys.Res 90: In: Rencv, A.N(Ed), Remote Sensing for the Earth Sciences: 797-804. [View Article] Manual of Remote Sensing. Wiley pp: 59-110. [View Article] 19. Goetz AFH, Rock BN, Rowan LC (1983) Remote sensing for 7. Bryant RG (1996) Validated linear mixture modeling of Landsat TM exploration: an overview. Economic Geology 78: 573–590. [View data for mapping evaporite minerals on a playa surface: Methods and Article] applications. Int J Remote Sens 17: 315-330. [View Article] 20. Shalaby MH, Bishta AZ, Roz ME, El Zalaky MA (2010) 8. Chapman JE, Rothery DA, Francis PW, Pontual A (1989) Remote Integration of geologic and remote sensing studies for the sensing of evaporite mineral zonation in salt flats (salars). Int J discovery of uranium mineralization in some granite plutons, Remote Sens 10: 245-255. [View Article] Eastern Desert, Egypt. Journal of King Abdulaziz University: 9. Crowley JK (1991) Visible and near-infrared (0.4-2.5 µm) Earth Sciences 21: 1–25. [View Article] reflectance spectra of playa evaporite minerals. J Geophys Res 21. Vincent RK (1997) Fundamentals of Geological and 96: 16 231-240. [View Article] Environmental Remote Sensing, Prentice Hall, Upper Saddle 10. Crowley JK (1993) Mapping playa evaporite minerals with River. [View Article] AVIRIS data: A first report from Death Valley, California.Remote 22. Hunt GR, Salisbury, JW, Lenhoff CJ (1973) Visible and near- Sens Environ 44: 337-56. [View Article] infrared spectra of minerals and rocks. IV. Sulfides and sulfates. 11. Crowley JK, Hook SJ (1996) Mapping playa evaporite minerals Mod Geol 4: 85-106. [View Article] and associated sediments in Death Valley, California, with 23. Lindberg JD, Smith, MS (1973) Reflectance spectra of gypsum multispectral thermal infrared images. J Geophys Res 101: 643- from the National Monument and basalt from 660. [View Article] nearby lava flow. Am Mineral 58: 1062-64. [View Article] 12. Öztan NS, Süzen ML (2011) Mapping evaporate minerals by 24. Clark RN, Swayze GA, Gallagher A, King TVV, Calvin WM Aster. International Journal of Remote Sensing 32: 1651-1673. (1993) The U.S. Geological Survey, Digital Spectral Library: [View Article] Version 1:0.2 to 3.0 µm. U.S. Geological Survey 93-592. [View 13. Fujisada H (1995) Design and performance of ASTER instrument. Article] SPIE 2583: 16-25. [View Article] 25. Davis DL (1993) Gypsum In: Minerals Yearbook U.S Bureau of 14. Yamaguchi Y, Anne BK, Hiroji T, Kawakami T, Pniel M (1998) Mines. Department of the Interior. [View Article] Overview of Advanced Spaceborne Thermal Emission and 26. Javor B (1989) Hypersaline Environments: Microbiology and Reflection Radiometer (ASTER).Geosci. Remote Sens 36: 1062- Biochemistry. Springer. [View Article] 71. [View Article]

Citation: Ghrefat HA (2019) Mapping Maqna Gypsum Deposits in the Midyan Area, Northwestern Saudi Arabia, Using ASTER Thermal Data. J Geol Geosci 4: 001-006. Copyright: © 2019 Ghrefat HA. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

J Geol Geosci 6 Volume 4(1): 2020