Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/taylor-francis/mapping-of-ferric-fe3-and-ferrous-fe2-iron-oxides-distribution-using-9OAnn1WA1s
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Taylor & Francis
- Copyright
- © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON).
- ISSN
- 2474-9508
- DOI
- 10.1080/24749508.2022.2130556
- Publisher site
- See Article on Publisher Site
Abstract
GEOLOGY, ECOLOGY, AND LANDSCAPES INWASCON https://doi.org/10.1080/24749508.2022.2130556 RESEARCH ARTICLE 3+ 2+ Mapping of ferric (Fe ) and ferrous (Fe ) iron oxides distribution using ASTER and Landsat 8 OLI data, in Negash Lateritic iron deposit, Northern Ethiopia a b b b Haylemikeal Hans Abay , Dagnachew Legesse , Karuturi Venkata Suryabhagavan and Balemwal Atnafu a b Mekelle University, Mekelle, Ethiopia; School of Earth Sciences, Addis Ababa University, Addis Ababa, Ethiopia ABSTRACT ARTICLE HISTORY Received 31 May 2022 Iron plays an important role in industrial and engineering fields development of a country and Accepted 26 September 2022 as such there is an enormous demand for iron in Ethiopia. However, a search for this valuable primary mineral resource exploration remains challenging and costly. Therefore, this study KEYWORDS aims to map iron oxide minerals using Landsat-8/operational land imager (OLI) and advanced ASTER; band ratio; space-borne thermal emission and reflection (ASTER) satellite imagery in Negash Lateritic iron endmember extraction; PCA; deposit, Northern Ethiopia to ease the costs and reduce the time. Different image processing iron oxides; LSU; MTMF techniques such as band ratio, selective principal component analysis, linear spectral unmixing, and mixture-tuned matched filter were used to produce iron oxide maps. Minimum noise fraction (MNF), pixel purity index (PPI), and N-dimensional visualizer were also applied to extract endmembers in the automated spectral hourglass wizard. In addition to this, the enhanced image thresholding and scatter plot were used to map the potential areas. Ferric iron oxide band ratio of ASTER mapped maximum area of 62.1 km followed by a laterite band ratio of ASTER covering 57.8 km . The result was validated using existing iron oxide polygons and the outcome obtained from selective PCA shows a strong match with the existing iron oxide polygons. The sub-pixel mapping techniques show poor accuracy in mapping goethite and hematite relative to the pixel level. Thus, it is evident from the results that ASTER mapped better than Landsat 8 OLI for band ratios of selective PCA, unmixing, MTMF, and mineralized areas while characterizing with limited fieldwork. 1. Introduction Satellite images are widely used to map geological Iron is the world’s most commonly used mineral and environmental features at different scales (Morais resource accounting for about 95% of the annual et al., 2012; Guha et al., 2019; Kumar et al., 2020; Shaik metal use. China is currently the world′s largest con- et al., 2021; Yazdi et al., 2018). Now-a-days, mineral sumer of iron ore and also the world’s largest steel- detection using remote-sensing techniques is important since it saves time and effort, unlike manual land sur- producing country followed by Japan and Korea veys and many satellite remote-sensing data sets are (Chen et al., 2020; Govil et al., 2018; Mohamed et al., accessible freely and could be extensively used for 2021; Ranjbar et al., 2004). Iron is primarily used in mineral exploration. Spectral absorption features are structural engineering, automobiles, and general often regarded as important tools for spatial mapping industrial applications. The northern part of Ethiopia of minerals/group of minerals specially associated with is endowed with a variety of minerals such as fossil hydrothermal deposits (Clark & Roush, 1984; Clark et fuel, metalliferous, and non-metalliferous minerals, al., 1995; Cloutis, 1996). Absorption features of miner- and laterite is a polymetallic area among them. als imprinted in their reflectance spectra as a result of Laterite is a consolidated product of humid tropical atomic processes operative within the ore. In recent weathering of a mix of goethite, hematite, kaolin, times, a few key economic rocks (chromite, kimberlite, quartz, some times bauxite, and other clay minerals, limestone, etc.) have also been delineated in spatial and appears as red or brown to chocolate colored at domain using space-borne/airborne sensors capable of the top with hollow, vesicular, and botryoidal struc- recording absorption features of these rocks (Guha et ture (Elsayed et al., 2020; Haldar, 2018; Schubert, al., 2014; Rajendran et al., 2012; El Zalaky et al., 2018). 2015). Iron deposits as laterite are also found in the In this regard, spectral features of different rocks are country at Wollega (Chago, Dha, Gordona-Korree, analysed in laboratories based on comparative analysis Worakalu, Belowtuist, Katta valley, Yubdo); Kaffw of reflectance spectra of constituent minerals within the (Garo, Melka Sedi, Dombova, Mai Guda); Sidamo visible-near infrared (VNIR) and shortwave infrared (Melka Arba), and Tigray (Adwa, Wukro and (SWIR) electromagnetic domains, and these Enticho) localities (Tadesse, 2009). CONTACT Karuturi Venkata Suryabhagavan drsuryabhagavan@gmail.com School of Earth Sciences, Addis Ababa University, Addis Ababa, Ethiopia © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON). This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2 H. H. ABAY ET AL. characteristics are related to certain chemical composi- 2011; Galvao et al., 2005; Guha et al., 2013; Hewson et tions and lattice structures of minerals and rocks (Ni et al., 2005; Hosseinjani & Tangestani, 2011; Kalinowski al., 2020). For this purpose, both VNIR and SWIR & Oliver, 2004; Mars & Rowan, 2010; Pal et al., 2011; regions of the spectrum encompassing 400–1000 nm Pour & Hashim, 2012; Rajendran et al., 2012; Van der and SWIR 1000–2400 nm, respectively, are measured Meer et al., 2012). (Transon et al., 2018). Unlike multispectral sensors as In Ethiopia, lack of modern exploration methods Landsat-8 (11 bands) that record relatively small num- such as space and airborne surveys in combination ber of discrete spectral bands (4–20), hyperspectral with ground reconnaissance to delineate promising sensors record a high number of continuous and nar- iron ore zones makes it expensive and time-consuming. row spectral bands of 5–15 nm (Kaufmann et al., 2009). Therefore, this study aims to precisely map the ferric 3+ 2+ This information is useful for the potential mapping of (Fe ) and ferrous (Fe ) iron oxides distribution using heavy metal contaminations and reolith characteristics ASTER and Landsat 8 OLI data for the first time in in the context of mineral deposits. Negash Lateritic iron deposit of Northern Ethiopia to Numerous examples of the application of spectral save on the costs and duration of exploration. remote-sensing techniques in exploring iron ores are available around the world (Azizi & Saibi, 2015; M. 2. Material and methods Azizi et al., 2015; Mogren et al., 2017; Saadi et al., 2008a, 2008b; Saibi et al., 2018). Bersi et al. (2016) 2.1 Study area have used a combination of remote-sensing and aero The study area is situated at the Eastern Tigray Northern gravity to evaluate the ore potential of Gara Djebilet, Ethiopia at four different woredas located at about 57 km southwestern Algeria and to estimate the tonnage of the from Mekelle city. Geographically, the area is located in iron ore at Gara Djebilet deposits. Saibi et al. (2018) zone 37 bounded by UTM coordinates of have reviewed the applications of remote-sensing in 557,000 − 578,000 m E, and 1,525,000 − 1,546,500 m N geosciences. Yazdi et al. (2018) have successfully applied covering a total area of 507.6 km (Figure 1). The locality alteration mapping for porphyry copper exploration is accessed by an asphalt road running from Mekelle to using ASTER and Quick Bird multispectral images. Adigrat and gravel road to Negash and Wukro besides Similar studies on hydrothermally altered mineral map- alternative small trail routes to other directions. ping were conducted by Nabilou et al. (2018), Fakhari et al. (2019), Zamyad et al. (2019). All these works indicate the credibility of using remote-sensing datasets as a 2.2 Geology cost-effective tool compared to geophysical and geo- According to Gebresilassie et al. (2012), the area forms a chemical techniques for mapping hydrothermally part of the Arabian Nubian Shield and consists of altered minerals. Neoproterozoic low-grade N-S to NE-SW trending base- Various researchers have reported the capability of ment rocks of Tsaliet Group (~860 − 750 Ma) with ASTER and Landsat 8 OLI sensors in image proces- metavolcanics, meta volcaniclastics, metasediments, and sing and image enhancement techniques such as younger Tambien Group (~740 Ma) with metasedi- Principle Component Analysis (PCA), Minimum ments, slate, phyllite, meta limestone, and pebbly slate Noise Fraction (MNF), Band Ratios (BRs), Band (diamictite). Apart from foliation, Tambien Group rocks Combinations (BCs), and spectral indices for iron show development of synclinal structures (Negash syn- oxide mapping (Fantaye, 2009; Gad & Kusky, 2006, cline). These structures are intruded by post-tectonic 2007; Omer & Elsayed Zeinelabdein, 2018; Rajendran granitoid (~600 Ma) and overlain unconformably by et al., 2012; Van der Meer et al., 2012). Spectral map- fluvial Paleozoic iron-rich Enticho Sandstone and Edaga ping algorithms such as Spectral Angle Mapper Arbi Tillite and by marine Mesozoic iron-rich Adigrat (SAM), Spectral Feature Fitting (SFF), Matched sandstone, Antalo simestone, Agula shale, and Amba Filter, Constrained Energy Minimization (CEM), Aradom sandstone. Dolerite dikes also have intruded Linear Spectral Unmixing (LSU), and Mixture Tuned during uplift and faulting during Cenozoic time. Tsaliet Matched Filter (MTMF) are well employed on ASTER metavolcanics cover the largest area (35.19%) followed by datasets to obtain the lithological, mineral, and hydro- meta-greywacke (15.28%; Figure 2). thermal alteration maps with reasonable accuracies (Boardman & Kruse, 2011; Gad & Kusky, 2007; Galvao et al., 2005; Gopinathan et al., 2020; 2.3 Data Hosseinjani & Tangestani, 2011; Pour & Hashim, 2012; Pour et al., 2011, 2018; Qiu et al., 2006; In this study, ASTER (Advanced Spaceborne Thermal Rajendran et al., 2013; Saed et al., 2022). Similar stu- Emission and Reflection Radiometer) and Landsat 8 dies using ASTER were carried out for geospatial and Operational Land Imager (OLI) image data covering geological mapping of prospective iron ore zones the suspected area of iron ore deposits in Negash were (Aboelkhair et al., 2011; H. Azizi et al., 2010; Bedini, used. In this effort, VNIR and SWIR wavelengths were GEOLOGY, ECOLOGY, AND LANDSCAPES 3 Figure 1. Location map of the study area. employed to decipher spatial distribution of iron oxide Spectral Hypercubes (FLAASH) tropical model and sites (Table 1). Landsat 8/Operational Land Imager rural aerosol model were adopted. On the other hand, images of six spectral bands (2 to 7) with a spatial Landsat-8/OLI DN values were converted to radiance resolution of 30 m were utilized in this study. using radiometric calibration, followed by transfer to Approximate scene size is 170 km north-south by BIL format and then applying FLAASH atmospheric 183 km east-west (106 mi by 114 mi). Secondary correction. Tropical atmospheric conversion was fol- data including a geological map of the Wukro sheet, lowed by band match to scale reflectance between 0 a topo-sheet of the Wukro area and shapefiles of and 1 and subset using study area. Finally, NDVI was different objects were supplemented. Geological sur- calculated for both the images. Areas greater than 0.4 vey and laboratory analysis were carried out to con- for ASTER and 0.3 for Landsat 8 OLI were masked as firm the image processing results and iron formations vegetation and ignored during further analysis prospectivity mapping them. A global positioning sys- (Tompolidi et al., 2020). The NDVI for both images tem (GPS) survey was also conducted in the study area was calculated from equation 1. Panchromatic and for verifying the spatial distribution of alteration zones thermal infrared (TIR) bands, as well as bands 1, 8, and lithological units using a handheld GPS. and 9 of the OLI were excluded from the analysis as Additionally, numerous photos were taken from bands 1 and 9 are meant for coastal studies and cirrus alteration zones and lithological units during the cloud detection, respectively (Masoumi et al., 2017). field work. Overall methodological framework and NDVI ¼ NIR Red=NIRþ Red (1) data analysis are presented in Figure 3. The remote- where NIR is band 3 and 5, and red is band 2, and 4, sensing datasets were processed using the ERDAS for ASTER and Landsat 8 OLI, respectively. Imagine and ArcGIS software packages. 2.4 Data preprocessing 2.5 Endmember extraction method Preprocessing applied to ASTER imagery consisted of Endmember extraction is an important process in the conversion from the DN value to radiance using radio- creation of useful material abundance maps. Although metric calibration, layer stacking, conversion to BIL there are different endmember extraction methods, format, and atmospheric correction. For atmospheric image spectra in this study were extracted through a correction, Fast Line-of-sight Atmospheric Analysis of “spectral endmember selection” procedure, including 4 H. H. ABAY ET AL. Figure 2. Geological map of Negash area. Table 1. Performance characteristics of satellite data. Data set Band name Band width (µm) Spatial resolution (m) Acquisition date Source ASTER 1 VIS 0.52 − 0.60 15 Dec 10, 2005 http://earthexplorer.usgs.gov/ 2 VIS 0.63 − 0.69 15 3 N NIR 0.78 − 0.86 15 4 SWIR 1.600 − 1.700 30 5 SWIR 2.145 − 2.185 30 6 SWIR 2.185 − 2.225 30 7 SWIR 2.235 − 2.285 30 8 SWIR 2.295 − 2.365 30 9 SWIR 2.360 − 2.430 30 Landsat Band 2 – Blue 0.45 − 0.51 30 28 January 2020 https://glovis.usgs.gov/ 8 OLI Band 3 – Green 0.53 − 0.59 30 Band 4 – Red 0.64 − 0.67 30 Band 5− Near infrared 0.85 − 0.88 30 Band 6− SWIR 1 1.57 − 1.65 30 Band 7− SWIR 2 2.11 − 2.29 30 minimum noise fraction (MNF), pixel purity index (Adams, J.B.J.R.G.A.E. & Composition, M, 1993; Shi & (PPI; Boardman, 1993; Boardman et al., 1995; Wang, 2016). Hosseinjani & Tangestani, 2011) and n-dimensional visualization (Boardman et al., 1995; Hosseinjani & 2.5.1 Pixel purity index (PPI) Tangestani, 2011). An automated spectral hourglass First, minimum noise fraction was applied for ASTER was used to run the steps sequentially and the image 9 bands and 6 Landsat 8 OLI bands. Then, Pixel Purity endmembers were obtained at the same spatial scale as Index (PPI) was run on the MNF data to aid in deriv- the image to be analyzed, whereas the reference end- ing endmembers from the image besides spatial data members were collected under different atmospheric reduction. All 9 ASTER and 6 Landsat 8 OLI were conditions than airborne or satellite imagery and at a used for further analysis since the eigenvalues are >1 different spatial scale due to their proximity to objects (Hosseinjani & Tangestani, 2011). The number of GEOLOGY, ECOLOGY, AND LANDSCAPES 5 Figure 3. Flowchart of the Methodology. iterations and thresholds used were 10,000 and 2.5 for (Hosseinjani & Tangestani, 2011). SFF values > 0.756 both images. were taken as endmember for ASTER image and 0.948 for Landsat 8 OLI (Kalinowski & Oliver, 2004). 2.5.2 N-Dimensional visualizer 2.5.3 Spectral feature fitting (SFE) The n-dimensional visualizer is an interactive tool that Spectral feature fitting was used to compare the fit of allows the user to select endmembers in n-space. Pixels image spectra to reference spectra using a least- from the spectral bands are loaded into an n-dimensional squares technique. Because SFF is an approach based scatter plot and rotated on the visualization tool until on absorption features after the continuum is elimi- points or extremities on the scatter plot are exposed. The nated from both datasets, the reference spectra are ENVI n-dimensional visualizer was loaded with the top- scaled to match the images spectra. scoring pixels from the PPI result and 10 and 7 of end- members were retrieved from n-dimensional visualizer automatically for ASTER and Landsat 8 OLI. These end- 2.6 Abundance mapping techniques members were used for subsequent classification and other processing. Then, using the spectral feature fitting, 2.6.1 Band ratio the known reference spectral was matched with an The band ratio is a simple and effective method for unknown spectral derived from the images identifying and demarcating iron ore mineral 6 H. H. ABAY ET AL. occurrences (Gopinathan et al., 2020). Band ratios 6/4 2.6.5 Correlation and model validation (ferrous iron oxides) and 4/3 (ferric iron oxides) of To correlate the results obtained from remote-sensing Landsat 8 OLI (Cardoso-Fernandes et al., 2019) and analysis Pearson correlation coefficient was used. To 3+ band 2/band 1 (for Fe ), 4/5 (Laterite) and band 5/ operate 1000 random points were generated using 50 2+ band 3 + band 1/band 2 (for Fe ) of ASTER were used rows and 50 columns, then correlation was calculated in this study (Gopinathan et al., 2020). from ASTER and Landsat 8 OLI. The Pearson correla- tion coefficient results were interpreted as strong cor- relation values ranging from ±0.50 to ± 1, a medium 2.6.2 Feature-oriented principal component correlation is defined as ± 0.30 to ± 0.49, and a weak selection (FPCS) correlation is defined as one with a value of less than ± Feature-oriented main component selection or Crosta 0.30 (Bekele et al., 2022). Technique is a method based on PCA (Traore et al., Validation of accuracy is a process that involves 2019). Only the bands of the image that have a reflec - evaluating the accuracy of a product and becomes an tion and absorption are used in feature-oriented prin- integral part of any remotely sensed data-derived map. cipal component selection. The ability to forecast It is critical to know the maps accuracy before making whether the target surface type is highlighted by dark any decisions based on it. The most frequent metric of or bright pixels in the matching principal component map accuracy is positional accuracy, which is a mea- image is a key feature of this method. In this investiga- sure of how closely the imagery matches with the tion, the Crosta approach was utilized and ASTER ground truth. Although there are different accuracy bands 1, 2, 3 and 4 (Traore et al., 2019) and Landsat assessment techniques, in this study, visual interpreta- 8 OLI bands 2, 4, 5 and 6 (Osinowo et al., 2021) were tion and existing iron oxide maps of the study area employed for iron oxide mapping. were digitized into polygons and overlain to the mapped results to check the positional accuracy as 2.6.3 Linear spectral unmixing (LSU) applied by Foody (2002). The LSU is a sub-pixel image processing algorithm, which was used to determine the abundance of the 2.6.6. Anomalous (potential) area detecting minerals in each pixel of an image. The reflectance at An anomaly is a pattern in the image data that each pixel of the image is assumed to be a linear does not follow the expected behavior, also referred combination of the reflectance of each material (end- to as outliers, exceptions, peculiarities (Chandola et member) present within the pixel. However, there are al., 2009; Zhou et al., 2016). To acquire quantitative certain limitations in applying the linear spectral information about the areas of mineral abundances, unmixing technique. The results of spectral unmixing thresholding (or density slicing) is applied to the are highly dependent on the input of endmembers and transformed data to separate the high potential changing endmembers also alter the final results pixels (anomalous areas) and exclude iron oxide (Gopinathan et al., 2020). In this study, LSU was lower concentrations as applied similarly by applied on the MNF images of the ASTER and (Wambo et al., 2020). A threshold was applied on Landsat 8 OLI images using different endmembers. band ratios, selected PCA and LSU which is gen- erated using the mean and standard deviation. The 2.6.4 Mixture tuned matched filtering (MTMF) standard deviation employed varies with the con- MTMF improves performance over MF alone by com- fidence level used to calculate the anomaly as bining components of hyperdimensional convex geome- applied by San et al. (2004). Before calculating the try spectral unmixing and statistically matched filter threshold, the results were stretched to 0 − 255 and target detection. It maintains the MF’s ease of calculation mean and standard deviation was taken from the while allowing for increased target selectivity, excelling at basic statistics. For MTMF results 2D scatter plot the accurate mapping of extremely small subpixel targets was used to identify the pixels with low infeasibil- with low false alarm rates. When utilized on the same ities and high MF scores. MF results in the x-axis input data as other methods that previously had a high and infeasibility in the y-axis were utilized. Pixels number of false alarms, MTMF frequently delivers low lying in the bottom right corner were selected as false alarm rates. This increase in performance demon- potential areas since they have high MF and low strates the effectiveness of the approach mixture tuning infeasibility values as done by Calin et al. (2015). (MT) component. More challenging applications are The threshold is calculated using equation 2. needed to fully quantify the detection versus false alarm rejection. MTMF was applied for iron oxide target detec- Threshold ¼ X þ SD (2) tion for both images using generated endmembers. Thresholding of MF score was used for target detection If confidence level is 92%, 2SD if 95% and 3SD if 98%, (Boardman & Kruse, 2011). where X is mean, and SD is standard deviation. GEOLOGY, ECOLOGY, AND LANDSCAPES 7 and a threshold value of 188 with 95% confidence 3. Results 2 2 level to an extent of 62.1 km and 13.2 km , respec- 3.1 NDVI tively for ASTER and Landsat 8 OLI (Table 2). The NDVI image of Landsat 8 OLI was in the range of −0.21 to 0.51 while that derived from ASTER lied 3.2.2 Laterite distribution between – 0.08 and 0.76 (Figure 4a and 4b). The Threshold values >210 with a 92% confidence level green portions show highly vegetated zones along were taken to prepare laterite distribution map using stream beds and valleys in contrast to less vegetated ASTER (Table 2). From Figure 6a, laterite distribution areas of moderate yellow regions and shadow and bare is seen dominated in the southwest and southeast. The lands of red patches. area covered by anomaly is 57.8 km . 3.2.3 Selective PCA distribution 3.2 Ratio maps Table 3 shows strong PCA4 values −0.79 and 0.60 of opposing signs between band 1 and 2 and a low con- 3.2.1 Ferrous and ferric iron tribution of bands 3 and 4. As a result, the target areas The ferrous mineral are abundance map generated were located around Graaras, Zarema, Adi mesanu from Landsat 8 OLI shows that the targets are in the and Abrha atsbha (Figure 6c). Similarly, Table 4 dis- Abrha-we-Atsbha area (Figure 5a). On the other hand, plays strong PCA4 values between 0.85 and −0.51 in ASTER map, the ore is found to be high in the Saze between bands 2 and 4 and a low contribution of area mountains and west of Tsenkanet (Figure 5b). bands 5 and 6. To map in bright the PCA4 was The anomalous area covered was 56.8 km and negated. As an outcome, the target areas were located 13.75 km for ASTER and Landsat images, respec- around Graaras and Abrha atsbha and Sendeda, tively. These areas were identified by a threshold Zarema and Adi mesanu (Figure 6b). value above 220 and confidence level of 92% in the case of ASTER and a threshold value of 197.2 and a confidence level of 95% in the instance of Landsat 8 3.3 End members extracted OLI (Table 2). Ferric iron identified through Landsat 8 is domi- Since the eigenvalues are >1 all 9 ASTER and 6 nant in the north around Sendeda, north of Graaras, Landsat 8 OLI were used in the analysis (Table 5). north of Genfel and southwest around Abrha Atsbha SFF values > 0.75 were taken as endmembers for areas (Figure 5c). Similarly, the demarks by ASTER are ASTER image (Table 6) and SFF values >0.95 as end- found in the cavity area around Adi Ksandid and other members for Landsat 8 OLI (Table 7). Finally, 5 for localities as Abraha Atsbha, Zarema and Graaras Landsat 8, and 8 for ASTER endmembers were used in (Figure 5d). The said areas were mapped using a this study (Figure 6d). The SFF values in ASTER were threshold value of 222.8 with 92% confidence level 0.85 and 0.81 for hematite and goethite, respectively, Figure 4. NDVI map generated from Landsat 8 OLI (a), and ASTER (b). 8 H. H. ABAY ET AL. Figure 5. Ferrous iron anomaly map generated from Landsat 8 OLI (a), ASTER (b), and ferric iron anomaly generated from Landsat 8 OLI (c), and ASTER (d). Table 2. Threshold, confidence level and area mapped using all techniques. Sensor Technique Confidence level Threshold used Area in (km ) ASTER Band 2/1 92% 222.8 62.1 Band 5/3 + 1/2 92% 220 56.8 Band 4/5 92% 210 57.8 Selective PCA 95% 210.5 34.79 Unmixing (goethite) 92% 197 8.4 Unmixing (hematite) 98% 185 26.5 Goethite MF 16.4 Hematite MF 52 Landsat Band 6/4 95% 197.2 13.75 8 OLI Band 4/3 95% 188 13.2 Selective PCA 95% 214.2 32.5 MF (goethite) 28.4 MF (hematite) 14.8 whereas they were 1 for both hematite and goethite in image. The bright areas indicate a high abundance of Landsat 8 OLI (Tables 6 and 7). the endmember whereas dark areas represent low abundance. The fraction image for the study site gives information about the relative abundance of 3.4 Linear spectral unmixing and mixture tuned the end member material considering each end mem- matched filtering ber present in a pixel. The analysis of MTMF produces a collection of rule images that correlate to both the The results of linear spectral unmixing are an abun- MF and infeasibility scores for each pixel when dance map for each end member and one RMS error GEOLOGY, ECOLOGY, AND LANDSCAPES 9 Figure 6. Laterite anomaly map generated from ASTER band ratio (a), anomaly map generated from PC4 Landsat 8 OLI (b), ASTER PC4 (c), and (d) final endmembers extracted from (a) ASTER, and (b) Landsat 8 OLI. Table 3. ASTER bands 1, 2, 3 and 4 eigenvector loadings. maximum of 26.4 km cover whereas goethite unmix- Eigenvectors Band 1 Band 2 Band 3 Band 4 ing has a least area 8.3 km . In Figures 7a and 7b, Saze PC 1 −0.332 −0.473 −0.515 −0.634 area shows high goethite and hematite anomalous area. PC 2 −0.428 −0.512 −0.152 0.729 From the MTMF results of both images, the potential PC 3 0.267 0.39 −0.844 0.254 PC 4 0.797 −0.602 −0.013 0.042 areas were identified based on MF scores and infeasi- bility images. Red and green areas in the scatter plots are abundance areas excluding false positives (Figure 7c compared to each endmember spectrum (two rule and 7d). The areas having high MF and low infeasibility images per endmember). Places with higher MF scores are shown in Figures 8a 8 8d. appear as brighter pixels in the MF images, showing areas with a large abundance of the associated endmember. 3.5 Correlation and model validation In order to map the anomalous areas from unmixing Landsat 8 OLI and ASTER for laterite results showed a results, a threshold value of 197 with 92% confidence high correlation (r = 0.59) in the case of selective PCA level for goethite and a threshold value of 188 with 98% and moderate correlation (r = 0.3) for ferric iron and a confidence level for hematite were chosen (Table 2). In poor correlation (r = 0.22) for ferrous iron. In terms of the area mapped, hematite unmixing has a Table 4. Landsat bands 2, 4, 5 and 6 eigenvector loadings. Eigenvectors Band 2 Band 4 Band 5 Band 6 PC 1 0.189 0.443 0.492 0.725 PC 2 −0.365 −0.468 −0.436 0.677 PC 3 −0.323 −0.567 0.753 −0.08 PC 4 0.853 −0.513 −0.010 0.098 10 H. H. ABAY ET AL. Table 5. MNF eigen values of ASTER and Landsat 8 OLI. Table 7. Unknown and reference spectral curves matching values of Landsat 8 OLI. Eigenvalues n-D class mean SFF Matched mineral ASTER Landsat 8 OLI 1 1 Goethite MNF 1 57.2 MNF 1 16.1 2 1 Kaolin MNF 2 29.3 MNF 2 7.8 3 1 Hematite MNF 3 11.1 MNF 3 6.4 4 1 Alunite MNF 4 8.2 MNF 4 5.8 5 0 No matching mineral MNF 5 6.2 MNF 5 3.4 6 1 Wollastonite MNF 6 5.8 MNF 6 1.99 7 0 No matching mineral MNF 7 3.8 MNF 8 3.3 MNF 9 2.6 negative correlation (r = −0.25; Figures 9a and 9b) made similar observations. Table 6. Unknown and reference spectral curve matching The ASTER results portray that the existing iron values of ASTER. oxide polygons overlay with the iron oxides of ferric n-D class mean SFF value Matched known mineral iron, laterite, and ferrous iron, but the band ratio 1 0 No matching mineral 2 0.95 Topaz obtained from Landsat 8 OLI was poor (Figure 10a). 3 0.81 Goethite Figure 10b obtained from selective PCA depicts a 4 0.96 Alunite 5 0.99 Topaz 2 better overlay with in the existing iron oxide polygons. 6 0 No matching mineral Although, both overlays point out similar results, 7 0.77 Lepidolite 8 0.85 Hematite selective PCA of ASTER fits better than the other 9 0.97 Pyrophyllite (Landsat 8 OLI). However, Figure 11 indicates that 10 0.82 Hornblende results obtained from unmixing and MTMF overlays fits better than the latter (ASTER). The results instances of unmixing and MTMF, both hematite and obtained from MF ASTER has a better fit with the goethite showed high positive (r = 0.5) and poor existing polygons. Figure 7. ASTER unmixing (a) Goethite anomaly and (b) Hematite anomaly, and scatter plots showing high MF and low infeasibility goethite and hematite of ASTER (c), and Landsat 8 OLI (d). GEOLOGY, ECOLOGY, AND LANDSCAPES 11 Figure 8. ASTER MTMF (a), Goethite anomaly and (b) Hematite anomaly and Landsat 8 OLI MTMF (c) Goethite anomaly and (d) Hematite anomaly. 4 Discussion vegetation interference helped to delineate the areas with high iron oxide content of values greater than 0.3 The comparative superior spectral and spatial charac- for Landsat 8 OLI and 0.4 for ASTER. A similar teristics of ASTER and Landsat 8 OLI images provide technique was employed by Tompolidi et al. (2020) useful information on the delineation of iron ore for mapping hydrothermal field in volcanic environ- deposits. In the wavelength ranges covered by ment in Greece. Vegetation masking in ASTER images Landsat TM bands 1 (blue) and 2 (green), which are was denser than that in Landsat 8 OLI images and analogous to bands 2 (blue) and 3 (green) in Landsat 8 these observations are in agreement with the findings OLI, vegetation and iron oxides show similar reflec - of Traore et al. (2019). tance spectra. In places with heavy or low vegetation Band ratio techniques were used to generate the cover, these bands are not very useful for differentiat - abundance of iron oxide content in various parts of ing iron oxides. In order to overcome this interference the study area. Each object has its own spectral reflec - while mapping the iron ore bands, first vegetation was tance pattern in different wavelength regions. The masked out during the present investigation. So, initi- object or a rock unit may have high reflectance value ally different vegetation indices such as ratio indices in some spectral regions, though it may be absorbed in and soil adjusted vegetation indices and NDVI were other spectral regions. For instance, iron ore absorbs mapped. In this method, since NDVI saturates above electromagnetic radiation (EMR) in the 0.85–0.9 μm 0.7 values, the maps were convenient in finding areas region. In this context, the band ratios serve as a that contain either the country rocks or the iron ore simple and powerful tool to identify and demarcate bands. This procedure was also adopted by the iron ore mineral deposits. As consequence, the Gopinathan et al. (2020) during mapping of iron oxi- results were obtained in different band ratios such as des in Tamil Nadu. The use of masking to remove 12 H. H. ABAY ET AL. Figure 9. Graphs showing correlation b/n results obtained from pixel-level image processing of Landsat 8 OLI and ASTER (a) and correlation b/n results obtained from Landsat OLI, and ASTER MTMF (b). 3+ band 2 and 1 for Ferric (Fe ), for band 5 and 3 + band ferrous iron map and laterite map generated from 2+ 1 and 2 for Ferrous iron (Fe ), and band 4 and 5 for ASTER could be explained by the alteration of biotite laterite in ASTER and in band 6 and 4 for Ferrous (Fe in granite to iron oxides as interpreted by Cardoso- + 3+ ), band 4 and 3 for ferric iron (Fe ) in Landsat 8 OLI. Fernandes et al. (2019). Although band 4 and 2 were also tested to map ferric Different bands in multispectral data were often iron from Landsat 8 OLI image(s), the out put was highly correlated and contained similar informa- very poor. To distinguish the minerals, threshold value tion. To map iron oxides, four bands were chosen for each resultant image was computed statistically by from each ASTER and Landsat 8 OLI images. The way of mean and standard deviation, as followed simi- bands were selected based on absorptive and reflec - larly by San et al. (2004) during his work on compar- tive characteristics of the mineral of interest. As ison of band ratioing and spectral indices methods for given in Table 4, PCA 4 showed strong values detecting alunite and kaolinite minerals using ASTER (0.853 and −0.513) of opposing signs between data in Biga region, Turkey. The values above the bands 2 and 4, but low contribution from bands 5 threshold were overlain on the hill shade images of and 6. Similarly, in Table 3, PCA 4 showed strong ASTER and Landsat 8 OLI. The high anomalous iron values (−0.79 and 0.60) of opposing signs between content in the western part of the study area in the band 1 and 2, but low contribution from bands 3 GEOLOGY, ECOLOGY, AND LANDSCAPES 13 Figure 10. Ratio of anomalous areas overlay on polygons of the existing iron oxides (a), and selective PCA results overlay with existing iron oxide polygons (b). and 4. From both images, PCA 4 was selected since by −1, as done by Traore et al. (2019). And the the values were high with opposite magnitude for target was sliced using 210.5 and 214.2 thresholds PCA 4, of ASTER in bands 1 and 2 and in Landsat for ASTER and Landsat 8 OLI, respectively to map 8 in bands 2 and 4. Negative values in reflective anomalous areas. bands indicate that iron oxides are dark. But to In this study, minimum noise fraction was applied show them in bright, the PCA 4 was multiplied for 9 bands of ASTER and 6 bands of Landsat 8 OLI 14 H. H. ABAY ET AL. Figure 11. Unmixing and MTMF results overlay on existing iron oxide polygons. for spectral data reduction before extracting the end- (2015). Both goethite anomaly derived from ASTER members. All 9 ASTER and 6 Landsat 8 OLI were used unmixing as well as from Landsat 8 OLI MTMF for further analysis since the eigenvalues were >1. The showed high targets in the western part of the study pixel purity index (PPI) was run on the MNF data to area and could be explained by the alteration of biotite aid in deriving endmembers from the image and to in granite to iron oxides as interpreted by Cardoso- achieve spatial data reduction. The number of itera- Fernandes et al. (2019). tions and thresholds used were 10,000 and 2.5 for both The MF images project areas with higher MF scores images as applied similarly by Hosseinjani and as brighter pixels, thus highlighting the areas with a Tangestani (2011). SFF values >0.75 were taken as large abundance of respective endmember. Perfectly endmember for ASTER image and 1 for Landsat 8 mapped pixels have an MF score above the back- OLI and this is greater than the threshold value of ground distribution, which has some noise-limited 0.5 selected by (Kalinowski & Oliver, 2004). As a spread around zero, and a low infeasibility value. To sequel, five endmembers for Landsat 8 and eight end- map the potential areas, a 2-D scatter plot was pre- members for ASTER were used in this work. The SFF pared using MF score and infeasibility. Pixels at the values were 0.85 and 0.81 for ASTER and 1 for Landsat right bottom show high MF scores and low infeasibil- 8. Spectral unmixing results are highly dependent on ity depicting potential areas. MTMF is not dependent the input endmembers and changing the endmembers on the input endmembers. Although it was con- would alter the results and this fact was proved by strained to 1, with a negative notation that is physically Gopinathan et al. (2020). meaningless, the same was interpreted by Calin et al. Linear spectral unmixing determines the relative (2015) as background noise. abundances of materials depicted by multispectral A strong correlation was obtained (r = 0.59) between imagery based on material spectral characteristics. the Landsat 8 OLI and ASTER results selective PCA for The bright areas indicate a high abundance of the laterite, moderate correlation (r = 0.3) for ferric iron and endmember whereas dark areas represent low abun- a poor correlation for ferrous iron (r = 0.22). For dance. Although the unmixing result was constrained unmixing and MTMF results, both hematite and to 1, the output contained a negative value, which is goethite have a poor positive and negative correlation physically meaningless. Such negative values are the with an r each of 0.15 and −0.25 respectively which was background noises that were obtained by Calin et al. similarly interpreted by Bekele et al. (2022) for LST. GEOLOGY, ECOLOGY, AND LANDSCAPES 15 While validating the results obtained through selective Azizi, M., Saibi, H., & Cooper, G. R. J. (2015). Mineral and structural mapping of the Aynak-Logar Valley (Eastern PCA, the best fit was obtained for the existing iron oxide Afghanistan) from hyperspectral remote sensing data and polygons and the band ratio showing a poor match in aeromagnetic data. Arabian Journal of Geosciences, 8(12), Landsat 8 OLI image. The results obtained from hema- 10911–10918. https://doi.org/10.1007/s12517-015-1993-2 tite unmixing and goethite ASTER MTMF have a better Azizi, H., Tarverdi, M. A., & Akbarpour, A. (2010). fit with the existing polygons whereas hematite ASTER Extraction of hydrothermal alterations from ASTER SWIR data from east Zanjan, northern Iran. Advances MF showed a good overlay. The MTMF obtained from in Space Research, 46(1), 99–109. https://doi.org/10.1016/ Landsat 8 OLI and goethite unmixing showed a very j.asr.2010.03.014 poor overlay. The comparison thus shows that ASTER Bedini, E. (2011). Mineral mapping in the kap simpson com- mapped better than Landsat 8 OLI for band ratios, plex, central East Greenland, using HyMap and ASTER selective PCA, unmixing and MTMF. remote sensing data. Advances in Space Research, 47(1),60– 73. https://doi.org/10.1016/j.asr.2010.08.021 Bekele, N. K., Hailu, B. T., & Suryabhagavan, K. V. (2022). Spatial patterns of urban blue-green landscapes on land Conclusion surface temperature: A case of Addis Ababa, Ethiopia. The geo-spatial techniques used to identify, map and Current Research in Environmental Sustainability, 4,100146. https://doi.org/10.1016/j.crsust.2022.100146 demarcate the iron ore deposits in the present instance Bersi, M., Saibi, H., & Chabou, M. C. (2016). Aerogravity could profitably be made use of to locate several other and remote sensing observations of an iron deposit in mineral resources in the country. The method Gara Djebilet, southwestern Algeria. Journal of African employed is an advanced technique that takes little Earth Sciences, 116,134–150. https://doi.org/10.1016/j. time, modest money and tiny effort to characterize jafrearsci.2016.01.004 Boardman, J. W. (1993). Automated spectral unmixing of and map large resources effectively in a country like AVIRIS data using convex geometry concepts, in Proc. Ethiopia with many constraints. Summ. 4th JPL Airborne Geosci. Workshop, vol 1, 1114, , vol 1, 1114, JPL Publication 93−26. Boardman, J. W., & Kruse, F. A. (2011). Analysis of imaging Acknowledgments spectrometer data using $N$-Dimensional geometry and a mixture-tuned matched filtering approach. IEEE We are thankful to the School of Earth Sciences, Addis Transactions on Geoscience and Remote Sensing, 49(11), Ababa University, for providing facilities, and funds. Mr. 4138–4152. https://doi.org/10.1109/TGRS.2011.2161585 Haylemikeal grateful to Mekelle University to sponsor for Boardman, J. W., Kruse, F. A., & Green, R. O. (1995). his higher studies in Remote-sensing and Geo-informatics. Mapping target signatures via partial unmixing of We are also gratefully acknowledge two anonymous AVIRIS data, in Proc. Summ. 5th Annu. JPL Airborne reviewers and Dr. Daren Jones, Associate Editor for their Earth Sci. Workshop, 23–26. critical and constructive comments of the manuscript. Calin, M. A., Coman, T., Parasca, S. V., Bercaru, N., Savastru, R. S., & Manea, D. (2015). Hyperspectral ima- ging-based wound analysis using mixture-tuned matched Disclosure statement filtering classification method. Journal of Biomedical Optics, 20(4), 046004. https://doi.org/10.1117/1.JBO.20. No potential conflict of interest was reported by the author(s). 4.046004 Cardoso-Fernandes, J., Teodoro, A. C., & Lima, A. (2019). Remote sensing data in lithium (Li) exploration: A new ORCID approach for the detection of Li-bearing pegmatites. International Journal of Applied Earth Observation and Karuturi Venkata Suryabhagavan http://orcid.org/0000- Geoinformation, 76, 10–25. https://doi.org/10.1016/j.jag. 0003-2528-9106 2018.11.001 Chandola, V., Banerjee, A., & Kumar, V. (2009). Anomaly detection: A survey. ACM Computing Surveys (CSUR), 41 References (3), 1–58. https://doi.org/10.1145/1541880.1541882 Chen, X., Mao, J., & Tian, H. J. S. (2020). Analysis of China’s Aboelkhair, H., Ninomiya, Y., Watanabe, Y., & Sato, I. iron trade Flow: Quantity, value and regional pattern. (2011). Processing and interpretation of ASTER TIR Sustainability, 12(24), 10427. https://doi.org/10.3390/ data for mapping of rare-metal-enriched albite granitoids su122410427 in the central eastern desert of Egypt. Journal of African Clark, R. N., & Roush, T. L. (1984). Reflectance spectro- Earth Sciences, 58,141–151. scopy: Quantitative analysis techniques for remote sen- Adams, J.B.J.R.G.A.E., & Composition, M. (1993). Imaging sing applications. Journal of Geophysical Research, 89, spectroscopy: (pp. 145–166). Interpretation based on spec- 6329–6340. tral mixture analysis. Clark, R. N., Swayze, G. A., Heidebrecht, K., Green, R. O., & Azizi, M., & Saibi, H. (2015). Integrating gravity data with Goetz, A. F. H. (1995). Calibration to surface reflectance of remotely sensed data for structural investigation of the terrestrial imaging spectrometry data: Comparison of Aynak-Logar Valley, eastern Afghanistan, and the sur- methods: Summaries of the fifth annual JPL airborne rounding area. IEEE J. Selected Top. IEEE Journal of Earth science workshop. JPL Publication. Selected Topics in Applied Earth Observations and Cloutis, E. A. (1996). Review article hyperspectral geological Remote Sensing, 8(2), 816–824. https://doi.org/10.1109/ remote sensing: Evaluation of analytical techniques. JSTARS.2014.2347375 16 H. H. ABAY ET AL. International Journal of Remote Sensing, 17(12), 2215– Kumaon Himalaya, India, using hyperspectral remote 2242. https://doi.org/10.1080/01431169608948770 sensing and hydrothermal alteration. Ore Geology de Morais, M. C., Martins, P. P., Júnior, & Paradella, W. R. Reviews, 92, 271–283. https://doi.org/10.1016/j.ore Multi-scale approach using remote sensing images to georev.2017.11.023 characterize the iron deposit N1 influence areas in Guha, A., Kumar, K. V., Rao, E. N. D., & Parveen, R. (2014). Carajás Mineral Province (Brazilian Amazon). (2012). An image processing approach for converging ASTER- Environmental Earth Sciences, 66(7), 2085–2096. 2012. derived spectral maps for mapping Kolhan limestone, https://doi.org/10.1007/s12665-011-1434-9 Jharkhand, India. Current Science, 106(1), 2–11. https:// doi.org/10.18520/CS/V106/I1/40-49 Elsayed, Z. K. A., El-Nadi, A. H. H., & Babiker, I. S. (2020). Guha, A., Singh, V. K., Parveen, R., Kumar, K. V., Prospecting for gold mineralization with the use of Jeyaseelan, A. T., & Dhanamjaya Rao, E. N. (2013). remote sensing and GIS technology in North Kordofan Analysis of ASTER data for mapping bauxite rich pockets State, central Sudan. Scientific African, 10, e00627. https:// within high altitude lateritic bauxite, Jharkhand, India. doi.org/10.1016/j.sciaf.2020.e00627 International Journal of Applied Earth Observation and El Zalaky, M. A., Essam, M. E., & El Arefy, R. A. (2018). Geoinformation: ITC Journal, 21,184–194. https://doi. Assessment of band ratios and feature-oriented principal org/10.1016/j.jag.2012.08.003 component selection (FPCS) techniques for iron oxides Guha, A., Yamaguchi, Y., Chatterjee, S., Rani, K., & Vinod mapping with relation to radioactivity using landsat 8 at Kumar, K. (2019). Emittance spectroscopy and broad- Bahariya Oasis. Egypt Res, 10(4), 1–10. https://doi.org/10. band thermal remote sensing applied to phosphorite 7537/marsrsj100418.01 and its utility in geoexploration: A study in the parts of EMD. (2011). (Ezana mining development), Mineral Rajasthan, India. Remote Sensing, 11(9), 1003. https://doi. resource estimate and technical report on iron deposit of org/10.3390/rs11091003 ShireMentebteb area, Northwestern of Tigray region, Haldar, S. K. (2018). Mineral exploration: Principles and Ethiopia. applications. Elsevier. https://doi.org/10.1016/C2017-0- Fakhari, S., Jafarirad, A., Afzal, P., & Lotfi, M. (2019). 00902-3 Delineation of hydrothermal alteration zones for por- Hewson, R. D., Cudahy, T. J., Mizuhiko, S., Ueda, K., & phyrysystems utilizing ASTER data in Jebal-Barez area, Mauger, A. J. (2005). Seamless geological map generation SE Iran. Iranian Journal of Earth Sciences, 11(1), 80–92. using ASTER in the Broken Hill-Curnamona province of https://doi.org/10.30495/ijes.2019.664780 Australia. Remote Sensing of Environment, 99(1-2),159– Fantaye, A. (2009). Mapping hydrothermally altered rocks 172. https://doi.org/10.1016/j.rse.2005.04.025 and lineament analysis through digital enhancement of Hosseinjani, M., & Tangestani, M. H. (2011). Mapping aster data case study: Kemashi area. Addis Ababa alteration minerals using sub-pixel unmixing of ASTER University. data in the Sarduiyeh area, SE Kerman, Iran. Foody, G. M. (2002). The role of soft classification techni- International Journal of Digital Earth, 4(6), 487–504. ques in the refinement of estimates of ground control https://doi.org/10.1080/17538947.2010.550937 point location. Photogrammetric Engineering and Kalinowski, A., & Oliver, S. (2004). ASTER mineral index Remote Sensing, 68(9), 897–904. http://eprints.soton.ac. processing manual. Remote Sensing Applications. uk/id/eprint/14895 Geoscience Australia, 37, 1–36. Gad, S., & Kusky, T. (2006). Lithological mapping in the Kaufmann, H., Segl, K., Guanter, L., Chabrillat, S., Hofer, S., Eastern Desert of Egypt, the Barramiya area, using Bach, H., Hostert, P., Müller, A., & Chlebek, C. (2009). Landsat thematic mapper (TM). Journal of African Review of EnMAP scientific potential and preparation Earth Sciences, 44(2), 196–202. https://doi.org/10.1016/j. phase. 6th EARSeL SIG IS Workshop. jafrearsci.2005.10.014 Kumar, C., Chatterjee, S., Oommen, T., & Guha, A. (2020). Gad, S., & Kusky, T. (2007). ASTER spectral ratioing for Automated lithological mapping by integrating spectral lithological mapping in the Arabian–Nubian shield, the enhancement techniques and machine learning algo- Neoproterozoic Wadi Kid area, Sinai, Egypt. Gondwana rithms using AVIRIS-NG hyperspectral data in Gold- Research, 11(3), 326–335. https://doi.org/10.1016/j.gr. bearing granite-greenstone rocks in Hutti, India. 2006.02.010 International Journal of Applied Earth Observation and Galvao, L. S., Formaggio, A. R., & Tisot, D. A. (2005). Geoinformation, 86, 102006. https://doi.org/10.1016/j.jag. Discrimination of sugarcane varieties in Southeastern 2019.102006 Brazil with EO-1 Hyperion data. Remote Sensing of Magendrana, T., & Sanjeevi, S. (2014). Hyperion image Environment, 94(4), 523–534. https://doi.org/10.1016/j. analysis and linear spectral unmixing to evaluate the rse.2004.11.012 grades of iron ores in parts of Noamundi, Eastern India. Gebresilassie, S., Bheemalingeswara, K., & Fiseha, A. (2012). International Journal of Applied Earth Observation and Geology and characteristics of meta-limestone-hosted Geoinformation, 26, 413–426. https://doi.org/10.1016/j. iron deposit near Negash, Tigray and northern Ethiopia. jag.2013.09.004 International Journal of Earth Sciences and Engineering, 5 Mars, J. C., & Rowan, L. C. (2010). Spectral assessment of (6), 1535–1544. new ASTER SWIR surface reflectance data products for Gopinathan, P., Parthiban, S., Magendran, T., Al-Quraishi, spectroscopic mapping of rocks and minerals. Remote A. M. F., Singh, A. K., & Singh, P. K. (2020). Mapping of 3+ 2+ Sensing of Environment, 114(9),2011–2025. https://doi. ferric (Fe ) and ferrous (Fe ) iron oxides distribution org/10.1016/j.rse.2010.04.008 using band ratio techniques with ASTER data and geo- Masoumi, F., Eslamkish, T., Honarmand, M., & Abkar, A. A. chemistry of Kanjamalai and Godumalai. Tamil Nadu, (2017). A comparative study of Landsat-7 and Landsat-8 South India, 18, 100306. https://doi.org/10.1016/j.rsase. data using image processing methods for hydrothermal 2020.100306 alteration mapping. Resource Geology, 67(1), 72–88. Govil, H., Gill, N., Rajendran, S., Santosh, M., & Kumar, S. https://doi.org/10.1111/rge.12117 (2018). Identification of new base metal mineralization in GEOLOGY, ECOLOGY, AND LANDSCAPES 17 Mogren, S., Saibi, H., Mukhopadhyay, M., Gottsmann, J., & Rajendran, S., Al-Khirbash, S., Pracejus, B., Nasir, S., Al-Abri, Ibrahim, E. (2017). Analyze the spatial distribution of lava A. H., Kusky, T. M., & Ghulam, A. (2012). ASTER detec- flows in Al-Ays Volcanic Area, Saudi Arabia, using tion of chromite bearing mineralized zones in Semail remote sensing. Arabian Journal of Geosciences, 10(6), Ophiolite Massifs of the northern Oman Mountains: 133. https://doi.org/10.1007/s12517-017-2889-0 Exploration strategy. Ore Geology Reviews, 44, 121–135. Mohamed, M. T. A., Al-Naimi, L. S., Mgbeojedo, T. I., & https://doi.org/10.1016/j.oregeorev.2011.09.010 Agoha, C. C. (2021). Geological mapping and mineral Rajendran, S., Nasir, S., Kusky, T. M., Ghulam, A., Gabr, S., & prospectivity using remote sensing and GIS in parts of El-Ghali, M. A. (2013). Detection of hydrothermal miner- Hamissana, Northeast Sudan. Journal of Petroleum alized zones associated with listwaenites in Central Oman Exploration and Production Technology, 11(3), 1123– using ASTER data. Ore Geology Reviews, 53, 470–488. 1138. https://doi.org/10.1007/s13202-021-01115-3 https://doi.org/10.1016/j.oregeorev.2013.02.008 Nabilou, M., Arian, M., Afzal, P., Adib, A., & Kazemi, A. Ranjbar, H., Honarmand, M., & Moezifar, Z. (2004). (2018). Determination of relationship between basement Application of the Crosta technique for porphyry copper faults and alteration zones in Bafq-Esfordi region, central alteration mapping, using ETM+ data in the southern Iran. Episodes Journal of International Geoscienc, 41(3), part of the Iranian volcanic sedimentary belt. Journal of 143–159. https://doi.org/10.18814/epiiugs/2018/018013 Asian Earth Sciences, 24(2), 237–243. https://doi.org/10. Ni, L., Xu, H., & Zhou, X. (2020). Mineral identification and 1016/j.jseaes.2003.11.001 mapping by synthesis of Hyperspectral VNIR/SWIR and Saadi, N., Aboud, E., Saibi, H., & Watanabe, K. (2008a). Multispectral TIR remotely sensed data with different Integrating data from remote sensing, geology and gravity classifiers, in IEEE. Journal of Selected Topics in Applied for geological investigation in the Tarhunah area, northwest Earth Observations and Remote Sensing, 13, 3155–3163. Libya. International Journal of Digital Earth, 1(4), 347–366. https://doi.org/10.1109/JSTARS.2020.2999057 https://doi.org/10.1080/17538940802435844 Omer, E. A. H., & Elsayed Zeinelabdein, K. A. (2018). Saadi, N., Watanabe, K., Imai, A., & Saibi, H. (2008b). Digital image processing of Landsat 8 and spectral analy- Integrating potential fields with remote sensing data for sis of ASTER data for mapping alteration minerals, geological investigations in the Eljufra area of Libya. Southern Hamisana, NE Sudan. Al Neelain. Journal of Earth, Planets and Space, 60(6), 539–547. https://doi. Geosciences , 2 (1), 10–20. http://hdl.handle.net/ org/10.1186/BF03353116 123456789/13555 . Saed, S., Azizi, H., Daneshvar, N., Peyman Afzal, P., Osinowo, O. O., Gomy, A., & Isseini, M. (2021). Mapping Whattam, S. A., & Mohammad, Y. O. (2022). hydrothermal alteration mineral deposits from Landsat 8 Hydrothermal alteration mapping using ASTER data, satellite data in Pala, Mayo Kebbi Region, Southwestern Takab-Baneh area, NW Iran: A key for further explora- Chad. Scientific African, 11, e00687. https://doi.org/10. tion of polymetal deposits. Geocarto International. 1016/j.sciaf.2020.e00687 https://doi.org/10.1080/10106049.2022.2059110 Pal, S. K., Majumdar, T. J., Bhattacharya, A. K., & Saibi, H., Bersi, M., Mia, M. B., Saadi, N. M., Al Bloushi, K. Bhattacharyya, R. (2011). Utilization of Landsat ETM+ M. S., & Akavian, R. W. (2018). Applications of remote data for mineral-occurrences mapping over Dalma and sensing in Geosciences. In M.-C. Hung & Y.-H. Wu Dhanjori, Jharkhand, India: An advanced spectral analy- (Eds.), Recent Advances and Applications in Remote sis approach. International Journal of Remote Sensing, 32 Sensing Book, Intech Edition (pp. 181–203). (Chapter 9). (14), 4023–4040. https://doi.org/10.1080/01431161.2010. San, B. T., Sumer, E. O., & Gurcay, B. (2004). Comparison of 484430 band ratioing and spectral indices methods for detecting Pour, A. B., & Hashim, M. (2012). The application of alunite and kaolinite minerals using ASTER data in Biga ASTER remote sensing data to porphyry copper and region, Turkey. In Proceedings ISPRS. epithermal gold deposits. Ore Geology Reviews, 44, 1–9. Schubert, G. (2015). Treatise on geophysics. Elsevier. https://doi.org/10.1016/j.oregeorev.2011.09.009 Settle, J. J. (2006). On the effect of variable endmember Pour, A. B., Hashim, M., Hong, J. K., & Park, Y. (2019). spectra in the linear mixture model, IEEE. IEEE Lithological and alteration mineral mapping in poorly Transactions on Geoscience and Remote Sensing, 44(2), exposed lithologies using Landsat-8 and ASTER satellite 389–396. https://doi.org/10.1109/TGRS.2005.860983 data: North-eastern Graham Land, Antarctic Peninsula, Shaik, I., Begum, S. K., Nagamani, P. V., & Kayet, N. (2021). Ore Geology Reviews, 108, 112–133. http://dx.doi.org/10. Characterization and mapping of hematite ore mineral 1016/j.oregeorev.2017.07.018 classes using hyperspectral remote sensing technique: A Pour, A. B., Hashim, M., & Marghany, M. (2011). Using case study from Bailadila iron ore mining region. SN spectral mapping techniques on short wave infrared Applied Sciences, 3(2), 182. https://doi.org/10.1007/ bands of ASTER remote sensing data for alteration s42452-021-04213-3 mineral mapping in SE Iran. International Journal of the Shi, C., & Wang, L. (2014). Incorporating spatial informa- Physical Sciences, 6(4), 917–929. https://doi.org/10.5897/ tion in spectral unmixing: A review. Remote Sensing of IJPS10.510 Environment, 149, 70–87. https://doi.org/10.1016/j.rse. Pour, A. B., Hashim, M., Park, Y., & Hong, J. K. (2018). 2014.03.034 Mapping alteration mineral zones and lithological units Shi, C., & Wang, L. (2016). Linear Spatial Spectral Mixture in Antarctic regions using spectral bands of ASTER Model. IEEE Transactions on Geoscience and Remote remote sensing data. Geocarto International, 33(12), Sensing, 54(6), 1–13. https://doi.org/10.1109/TGRS.2016. 1281–1306. https://doi.org/10.1080/10106049.2017. 2520399 1347207 Tadesse, S. (2009). Mineral resources potential of Ethiopia. Qiu, F., Abdelsalam, M., & Thakkar, P. (2006). Spectral Addis Ababa University Press. analysis of ASTER data covering part of the Tompolidi, A.-M., Sykioti, O., Koutroumbas, K., & Parcharidis, Neoproterozoic Allaqi-Heiani suture, Southern Egypt. I. (2020). Spectral unmixing for mapping a hydrothermal Journal of African Earth Sciences, 44(2), 169–180. field in a volcanic environment applied on ASTER, Landsat- https://doi.org/10.1016/j.jafrearsci.2005.10.009 8/OLI, and Sentinel-2 MSI satellite multispectral data: The 18 H. H. ABAY ET AL. Nisyros (Greece) case study. Remote Sensing, 12(24), 4180. Cameroon. Ore Geology Reviews, 122, 103530. https:// https://doi.org/10.3390/rs12244180 doi.org/10.1016/j.oregeorev.2020.103530 Transon, J. D., Andrimont, R., Maugnard, A., & Defourny, Winter, M. E. (2000). Comparison of approaches for deter- P. (2018). Survey of hyperspectral earth observation mining end-members in hyperspectral data. IEEE applications from space in the sentinel-2 context. Aerospace Conference. Proceedings (Cat. No. 00TH8484): Remote Sensing, 10(3), 157. https://doi.org/10.3390/ IEEE, 305–313. rs10020157 Yazdi, Z., Rad, J. A., Aghazadeh, M., & Afzal, P. (2018). Traore, M., Çan, T., & Tekin, S. (2019). Discrimination of Alteration mapping for porphyry copper exploration iron deposits using feature-oriented principal component using ASTERand QuickBird multispectral images, sona- selection and band ratio methods: Eastern Taurus/Turkey. jeel prospect, NW Iran. Journal of the Indian Society of International Symposium on Applied Geoinformatics, Remote Sensing, 46(10), 1581–1593. https://doi.org/10. ISAG2019 was held in Istanbul on, 9. 1007/s12524-018-0811-1 Van der Meer, F. D., Van der Werff, H. M., Van Ruitenbeek, Zamyad, M., Afzal, P., Pourkermani, M., Nouri, R., & Jafari, F. J., Hecker, C. A., Bakker, W. H., Noomen, M. F., Van M. R. (2019). Determination of hydrothermal alteration Der Meijde, M., Carranza, E. J. M., De Smeth, J. B., & zones using remote sensing methods in Tirka area, Woldai, T. (2012). Multi-and hyperspectral geologic Toroud, NE Iran. J Indian Soc Rem. Sens.Arab. J. remote sensing: A review. International Journal of Geosci.IEEE J. Selected Top. Appl. Earth Observ. Applied Earth Observation and Geoinformation, 14(1), Arabian Journal of Geosciences, 47, 1817–1830. https:// 112–128. https://doi.org/10.1016/j.jag.2011.08.002 doi.org/10.1007/s12524-019-01032-3 Wambo, J. D. T., Pour, A. B., Ganno, S., Asimow, P. D., Zhou, Z. G., Tang, P., & Zhou, M. (2016). Detecting anom- Zoheir, B., Dos Reis Salles, R., Nzenti, J. P., Pradhan, B., & aly regions in satellite image time series based on seasonal Muslim, A. M. (2020). Identifying high potential zones of autocorrelation analysis. ISPRS Annals of the gold mineralization in a sub-tropical region using Photogrammetry. Remote Sensing and Spatial Landsat-8 and ASTER remote sensing data: A case Information Sciences, 3, 303. https://doi.org/10.5194/ study of the Ngoura-Colomines goldfield, eastern isprs-annals-III-3-303-2016
Journal
Geology Ecology and Landscapes – Taylor & Francis
Published: Oct 14, 2022
Keywords: ASTER; band ratio; endmember extraction; PCA; iron oxides; LSU; MTMF