Skip to content
BY-NC-ND 4.0 license Open Access Published by De Gruyter Open Access October 3, 2017

Spectral properties of weathered and fresh rock surfaces in the Xiemisitai metallogenic belt, NW Xinjiang, China

  • Ke-Fa Zhou EMAIL logo and Shan-Shan Wang
From the journal Open Geosciences

Abstract

Surfaces weathering of rocks in which mineral materials may be similar to or quite different from the minerals in the underlying parent rock completely control the reflectance spectra of the terrain. Our study of typical weathered and fresh rock samples from the Xiemisitai metallogenic belt, Western Junggar region, Xinjiang, found that weathering results in the formation of new materials that cause differences in the spectral features of fresh and weathered rock surfaces. Alterations induce variations in spectrum brightness, presence and intensity of characteristic absorption features, and spectral slope. Spectral differences between weathered and fresh rock surfaces are small for rhyolite, granite, and tuffaceous sandstone, but large for andesite, basalt, and diorite. Spectral changes in the 350–1000 nm wavelength region are attributed to alteration of iron oxides by atmospheric processes or secondary alteration of iron-rich minerals. Spectral features between 1000–2500 nm are caused by O–H vibrations, with features at 2200–2500 nm solely attributed to hydroxyl groups. The strongest Al–OH bands appear near 2200 nm, while Mg–OH bands are found near 2300 nm and 2350 nm. Results from this study can be used to better characterize and discriminate lithological units and potential mineral zones using hyperspectral and multispectral remote sensing techniques.

1 Introduction

The ability to produce lithological maps for mineral exploration is one of the most useful contributions of hyperspectral imaging to the field of geology [1]. In arid and semi-arid environments, where vegetation cover is sparse and rock surfaces are exposed, ground-based geologic mapping is time-consuming, expensive, difficult, and even dangerous. In contrast, geologic mapping by remote sensing is economical, practical, and extensively used. Spectral properties of rocks mainly depend on their mineralogical composition, which produces characteristic absorption features in different wavelength regions [2]. Hence, absorption spectra datasets provide vital information about the diagnostic spectral features of lithological units and mineral assemblages [3].

Spectroscopic in situ and laboratory measurements are important components of any multispectral or hyperspectral remote sensing analysis [4, 5]. There are several important spectral library projects that have been established recently in this area of research such as the U.S. Geological Survey (USGS) spectral library [6], the Arizona State University (ASU), Thermal Emission Spectral Library [7], the ASTER spectral library [8], Specchio [4], the DLR spectral archive [9], and the globally distributed soil spectral library ICRAF–ISRIC (World Agroforestry Centre–International Soil Reference and Information Centre) [10]. These spectral libraries provide a large amount of standard spectral data and remote sensing image extraction of spectral endmembers. However, with hyperspectral remote sensing becoming more accurate, quantifiable, and readily available, the possibility of increasingly false identifications using standard spectral data from spectral libraries arises. Several reasons can be attributed to this high rate of false identifications: (1) many of these libraries contain laboratory measurements that use standardized acquisition procedures and well determined standards based on pure materials, but such pure materials are rare in natural settings; (2) most rock samples are analyzed using powders of varying particle size, and the mixture of these powders can mask diagnostic spectral features of rocks of interest and hamper their identification; and (3) most importantly, the weathering process produces mineral coatings that may cover the original parent rock surface, which means that the spectral signatures are derived from weathered surfaces [1113]. Therefore, standard spectral data of rocks are convenient, but are less useful for practical applications and accurate mineral mapping.

The majority of rocks exposed on the Earth’s surface are affected by weathering to some extent, depending on several factors including climatic zone, surface dynamics, and biological activity [1416]. Even on Mars, weathering rinds and coatings on rocky and sandy surface materials have developed over time, and are attributed to mineralogical and textural characteristics, duration of surface exposure, availability of water, or exposure to acidic volatiles [2, 17].

The two different types of rock coatings include biogenic (e.g., lichens, which consist mostly of mosses and cyanobacteria and form in humid environments [18, 19]) and non-biogenic (e.g., desert varnish) coatings. Desert varnish is a mixed layer of illite clays and nanocrystalline iron and manganese oxides, which partially cover rock surfaces in arid and semi-arid environments [2022]. Such rock coatings are usually only microns to millimeters in thickness, but they completely control the spectral signatures of the rock surfaces. Spectraldata from such surfaces must be used with caution due to possible misinterpretation of the rock’s mineralogical composition and lithological identification. However, identification and discrimination of these surface materials are of crucial importance for geological studies that use remote sensing to explore the mineralogical composition of the parent rocks [13].

In arid and semi-arid environments, weathering processes alter the shape, overall reflectance (brightness), and presence of certain characteristic absorption features of different rock types. Thus, it is necessary to study spectral differences between exposed and fresh surfaces in the field. However, to date there is very little literature that thoroughly describes the spectral characteristics of weathered and fresh rock surfaces in arid and semi-arid environments. The correlation of remotely sensed imagery to actual surface geology requires solid knowledge of the surficial characteristics of exposed rock outcrops. However, it is not always possible to find fresh surfaces in the field. One must differentiate upper surface spectra (and lower surface spectra) from the spectra of a rock’s interior [23, 24].

To correctly assess and interpret spectral data of rock surfaces, a better understanding of the changes occurring in the spectra of weathered and fresh rock is needed. Although several remote-sensing-related studies, including in situ spectral measurements in arid regions have been published [2528], none of those studies focused on the description of the spectral properties of surface materials in the Xinjiang region, China. Therefore, the primary aim of this paper is to report the spectral characteristics of the lithological units in West Junggar, Xinjiang, and to explore the spectral features of the most common minerals associated with weathered and fresh rocks at wavelengths from 350 nm to 2500 nm. We have selected this spectral range because of its use in many current airborne hyperspectral remote-sensing systems.

2 Data and Experimental Approach

2.1 Geological Setting

The West Junggar region of Xinjiang, located between Altay Shan and Tianshan, extends westward to the Junggar-Balkhash system adjacent to Kazakhstan, and eastward to the Junggar Basin in China (Figure 1a) [29]. The West Junggar tectonic unit consists of several terranes such as island arcs and accretionary complexes, including, from north to south, the Zharma-Saur arc, the Tarbagatay accretionary complex, the Chingiz arc, and the West Junggar accretionary complex (Figure 1b) [30, 31].

Figure 1 (a) Simplified schematic geological map of China and adjacent areas (modified by 29); (b) Generalized geological map of the northwestern Junggar region (modified by 29): (c) Geological map of Xiemisitai area showing the location of main rock samples.
Figure 1

(a) Simplified schematic geological map of China and adjacent areas (modified by 29); (b) Generalized geological map of the northwestern Junggar region (modified by 29): (c) Geological map of Xiemisitai area showing the location of main rock samples.

Xiemisitai Mountain, which belongs to the Chingiz arc, is one of several E–W-striking mountains in the northern part of West Junggar, and is bounded by two major faults, the northern Hong Gullah fault and the southern Xiemisitai fault [29]. In this region, Silurian island-arc-related felsic volcanic lavas and tuff, as well as granitic intrusions are widespread, and several newly discovered mineral deposits in the region are related to this magmatic event (Figure 1c). The Xiemisitai copper deposit is a volcanic-related porphyry copper deposit controlled by a caldera fracture system, which is superimposed on E–W-striking regional faults [28, 32].

2.2 Sample Analyses

Two surfaces of all lithological units were analyzed: a fresh surface obtained by sectioning the sample, and a weathered surface sampled in the field. Fresh surface analyses were conducted on rock fragments with little or no visible signs of surface alteration. The same samples were also used for thin section analyses and subsequent petrographic description. Weathered surfaces were analyzed as exposed in the field or on weathered parts of rock hand samples. Mineral assemblages of selected samples were confirmed by microscopic study, portable near-infrared mineral analyzer (BJKF-II) and X-ray fluorescence (XRF). For fresh sample surfaces, mineral composition was determined using plane-polarized and cross-polarized transmitted light and reflected light petrography with a Carl Zeiss Axio Scope A1 microscope. Maximum and minimum magnification were 50x × 10x and 2.5x × 10x, respectively. Mineral composition of the weathered surfaces was determined by BJKF-II portable near-infrared mineral analyzer (Nanjing Instrument Co., Ltd., China). BJKF-II is a portable instrument specially designed for field mineral exploration. The instrument adopts diffuse reflectance spectroscopy, scanning by grating spectrophotometer, integrating sphere detection, and microcomputer for spectral data processing to guide field mineral exploration [33].

Spectral data were acquired with a spectroradiometer (FieldSpec 4 Hi-Res, ASD, USA). The band extent varied from 350 nm to 2500 nm, and the spectral resolution was 3 nm and 8 nm at band ranges of 350–1000 nm and 1000–2500 nm, respectively. Samples were collected from the West Junggar and were measured at the laboratory in Urumqi. Spectral measurements were made using a fiber optic (FOV 25°) along the direction normal to the surface at a distance of 1 cm, which results in an approximately circular footprint of 0.45 cm on the sample surface. A100-W Reflecta halogen lamp mounted at an angle of 45° was used as an illumination source. The reflectance spectrum was obtained by calculating the ratio of the radiance of the sample to the radiance of a 99% reflective white reference panel (Spectralon, Labsphere, North Sutton, NH, USA) under the same illumination and viewing conditions. The spectrum recorded for each surface location is based on an average of 10 scans [5, 19].

Qualitative mineralogical analysis of rock surfaces was carried out using a hand-held XRF device Thermo Scientific NITON XL3t950, XRF, USA). Simultaneous analyses of up to 25 elements in the analytical range between sulphur (atomic number 16) and uranium (atomic number 92), as well as of light elements (Mg, Al, Si, P, S and Cl) were carried out. All analyses were performed in bulk mode (“mining mode”). Measement time was 120 seconds, using four different energy settings: light (Mg, Al, Si, P, S, Cl), low (Ca, K, Sc, Ti), main (K-lines from V to Ag and L-lines from Au to Pb), and high (Ag to Au). Analytical conditions included a maximum excitation source beam current of 80μA, acceleration voltage of 45 kV, and a beam diameter of 10 mm. The XRF is factory “calibrated” with a Fundamental Parameter algorithm-based program. As this software is propriety software of the Niton company, it is not possible to discern how it compensates for variable sample thicknesses, irregular-shaped surfaces or other matrix and geometric effects. In addition to the analyzed elements, the XRF software also estimates the non-measurable part of each spectrum, named “balance” (“Bal”). The balance is composed of the oxide portion of the compounds, organic matter, and – if present – carbonate and water.

2.3 Description of Sampled Materials

We paid close attention to the distribution of the samples (Table 1 and Figure 2), making sure that most of the lithological units, their weathering rinds, and rock coatings would be included. Some of the major lithological units were not considered due to difficulties in obtaining representative fresh and weathered rock samples. These limitations are due to the friable nature of sediments such as marls and conglomerates belonging to the Quaternary. The macroscopic description of mineral composition and the type of metamorphic and/or hydrothermal alteration (if present) of the lithological units selected for this study is given below.

Figure 2 Hand specimens and microscopic photographs, (a)(b): rhyolite, (c)(d): andesite, (e)(f): granite, (g)(h): Tuffaceous sandstone, (i)(j): basalt, and (k)(l): diorite; (Qtz – quartz, Pl – lagioclase, Hbl – hornblende, Tuff – tuffaceous). (b): plane–polarized light, 10–fold magnification; (f): plane–polarized light, 10–fold magnification; (l): plane–polarized light, 10–fold magnification; (d): cross–polarized light, 10–fold magnification; (h): cross–polarized light, 10–fold magnification and (j): cross – polarized light, 10–fold magnification.
Figure 2

Hand specimens and microscopic photographs, (a)(b): rhyolite, (c)(d): andesite, (e)(f): granite, (g)(h): Tuffaceous sandstone, (i)(j): basalt, and (k)(l): diorite; (Qtz – quartz, Pl – lagioclase, Hbl – hornblende, Tuff – tuffaceous). (b): plane–polarized light, 10–fold magnification; (f): plane–polarized light, 10–fold magnification; (l): plane–polarized light, 10–fold magnification; (d): cross–polarized light, 10–fold magnification; (h): cross–polarized light, 10–fold magnification and (j): cross – polarized light, 10–fold magnification.

Table 1

Mineralogical and textural description of the selected samples, (Qtz – quartz, Pl – plagioclase, Hbl – hornblende, Fsp – feldspar, Afs – alkalifeldspar, Cpx – clinopyroxene, Kln – kaolinite, Bt – biotite, Mag – magnetite, Chl – chlorite, Ep – epidote, Cal – calcite, Dol – dolomi, Ilm – ilmenite).

Sample No.LithologyTextureFresh surface (microscopic study)Weathered surface (BJKF-II)
XM01rhyoliteporphyriticQtz, Afs, Pl, glassQtz, Fsp, Kln
XM41andesiteporphyriticPl, Cpx, Hbl, Bt, glassBt, Mag, Fsp, Kln
XM12granitegranular and phaneriticQtz, Afs, Pl, BtQtz, Fsp, Kln, Chl, Ep
XM78tuffaceous sandstonehypautomorphic granularQtz, Pl, clastQtz, Cal, Dol, and iron oxides
XM04basaltpalimpsestCpx, Pl, glassFsp, Kln, Mag, Ilm,
XM20dioritehypidiomorphic granularHbl, Pl, BtPl, Kln, Bt, Mag, Ilm, and sulfides

Eighty rock samples were collected from bedrock outcrops. Six representative samples from the dominant lithological units in the study area include rhyolite, andesite, granite, tuffaceous sandstone, basalt and diorite. A summary of the sampled lithological units with their respective petrographic description is given in Table 1.

Rhyolite is a felsic volcanic rock, which is brick–red in the Xiemisitai area. It contains 70–80% quartz phenocrysts, 5-10% plagioclase, 10-25% glass, and is characterized by a porphyritic texture (Table 1 and Figure 1a, 1b). Feldspar and volcanic glass show signs of sericitization.

Andesite is an intermediate extrusive rock, often showing porphyritic texture. It is composed of phenocrysts (~20%), many of which contain biotite (~50%), hornblende (~20%), plagioclase (~30%), and glass (~80%) (Table 1, Figure 1c and 1d). Magnetite occurs along the borders of the phenocrysts and is altered slightly in the remainder of the samples.

The weathered surfaces of granite are largely pink–brown in color, with visible minerals beneath the surface. Under the microscope, the major mineral composition of the granite samples is identified as quartz (~40%), plagioclase (~20%), potassium feldspar (~30%), and biotite (~10%) with a distinct foliated fabric (Table 1, Figure 1e and 1f). The feldspar minerals show signs of sericitization and alteration to clay minerals, while biotites are characterized by chloritization.

Tuffaceous sandstone samples are medium -grained–poorly sorted, and are dominated by sub angular grains of feldspar (~10%), quartz (~10%), and a matrix (~80%) of volcanic glass. The tuffaceous sandstone in the Xiemisitai area is black, and can be categorized as lithic arkose tuff with cements consisting of tuffaceous material (Table 1, Figure 1g and 1h). The tuffaceous sandstone is strongly weathered, and most of the surface is covered with a layer of loess.

Basalt in the area is extremely fine-grained (<100 µm) and exhibits a uniform gray color on fresh surfaces. The basalt samples are composed of plagioclase (20–30%) and matrix (70–80%) (Table 1, Figure 1i and 1j). Minor fractures are partially filled with quartz with some minor rust staining on the edges. Weathered surfaces are characterized by a mixture of shades of gray or brown with a mesh of fine veins of quartz and calcite. Visible alteration of the interior of samples is limited.

Diorite is an intrusive igneous rock, mainly composed of quartz phenocrysts (20%), plagioclase (60-65%), and melanocratic minerals (15-20%). The diorite samples are grey to dark-grey in color with a greenish cast, but black or bluish-grey varieties also occur in the Xiemisitai area (Table 1, Figure 1k and 1l). The melanocratic minerals are altered into magnetite and chlorite.

All rock samples were analyzed for 25 elements (i.e., Ag, Al, Ba, Ca, Cd, Cr, Cs, Cu, Fe, K, Mg, Mn, Ni, Rb, S, Sb, Si, Sn, Sr, Te, Ti, V, Zn, Zr, and Bal) by XRF analysis (Table 2). The XRF analysis of weathered and fresh rock surfaces show that Si is the dominant element (>100000 ppm) followed by Al > Ca > Fe > K > Mg (>10,000 ppm). Elements with concentrations > 100 ppm are Mn > Ba > Sr > Te > Zr > Cs > V > Cu > Ni (>100 ppm). In general, element concentrations of weathered surfaces are similar to those of fresh surfaces.

Table 2

XRF measurements of six representative samples include rhyolite, andesite, granite, tuffaceous sandstone, basalt, and diorite. All measurements were taken in mining mode (120 seconds), and values for elements as PPM. Note: Totals are more than 100% as the balance (Bal; non measurable components) includes the oxide part of the compounds (O2 in SiO2 etc.); LOD: limit of detection of elementals; N/A: not applicable.

Lithology Sample NO.Rhyolite XM01Andesite XM41Granite XM12Tuffaceous sandstone XM78Basalt XM04Diorite XM20
weatheredfreshweatheredfreshweatheredfreshweatheredfreshweatheredfreshweatheredfresh
Bal(ppm)496539.34642294.75584266.06599218.00510439.44557253.44442635.56554706.44505032.13491037.19556025.19502678.56
Error2196.671863.281765.251955.432467.952220.952760.172043.8012237.672185.292127.292390.94
Ag(ppm)0.0013.02< LOD10.879.9211.3811.1211.30< LOD< LOD17.908.40
Error5.324.81N/A4.174.984.824.564.46N/AN/A5.164.24
Al(ppm)48382.3248414.1530555.6350646.5574487.6447194.8797506.9640771.3668179.9957599.3338533.2679545.63
Error1519.311410.791518.931412.482252.161791.042469.151469.733172.141739.042173.492116.18
Ba(ppm)337.17471.93602.75645.88717.50718.811162.68849.551119.23468.08932.66633.83
Error23.2330.1342.0927.5333.6232.2033.4630.7764.2341.1834.6828.37
Ca(ppm)2326.913155.68135823.256828.6348363.6933166.4623050.425527.823493.5021446.16250877.8948421.41
Error75.0076.121110.40118.61795.93710.81586.30131.19426.25529.501637.82668.82
Cd(ppm)12.6422.54< LOD22.8824.1733.2127.1322.84< LOD< LOD38.4119.39
Error6.498.34N/A7.338.838.658.067.82N/AN/A9.027.50
Cr(ppm)43.64145.7571.2016.56117.17185.48118.5742.24126.4254.07< LOD75.61
Error8.168.1922.388.6413.9730.2727.8523.0027.5927.09N/A21.00
Cs(ppm)89.97131.12< LOD120.34144.96136.61134.98137.95< LOD< LOD179.46121.73
Error6.828.80N/A7.639.338.888.408.24N/AN/A9.397.90
Cu(ppm)31.8752.3320.5417.0869.1781.6768.80399.9639.7053.43261.27168.13
Error9.9611.229.629.1513.1413.2513.0019.6911.1011.5217.7514.76
Fe(ppm)13460.4310699.3422193.5312811.3734060.0137310.4255366.0625631.3335416.0540016.6115051.5218803.28
Error202.85180.42247.59192.82433.02467.42465.92282.441811.21345.56226.47244.77
K(ppm)17593.6025137.4212206.7918385.929349.0012016.3318469.3821746.5457236.178099.7210742.8811627.34
Error251.37335.81257.97283.54250.29316.41367.41353.991766.55245.06327.85244.07
Mg(ppm)< LOD< LOD< LOD7619.07< LOD< LOD20840.0120440.0113127.887848.788251.6712344.86
ErrorN/AN/AN/A2308.26N/AN/A3696.164134.893581.782742.785266.673143.50
Mn(ppm)301.52215.071948.711243.42644.47783.011310.88209.222915.011723.45242.24356.93
Error44.4142.8381.5666.0465.3767.1281.5644.26102.2182.9145.8548.82
Ni(ppm)75.3561.8184.9262.99154.27139.5872.0993.84130.07114.53101.48115.82
Error21.0621.4021.2120.0126.9926.0524.9322.3623.4423.8123.6223.07
Rb(ppm)36.0647.1821.9635.5422.4814.3519.6737.6979.7918.4417.6721.34
Error1.431.621.071.381.271.041.231.483.572.081.111.18
S(ppm)920.791296.06646.90929.03480.49696.70896.0614709.992945.90640.571591.791634.97
Error239.9590.5282.6674.3079.8985.2480.50239.38151.3273.84125.9496.44
Sb(ppm)32.6078.50< LOD51.7560.9767.4954.2563.64< LOD< LOD87.8758.64
Error7.6410.21N/A8.6310.5110.169.449.36N/AN/A10.759.00
Si(ppm)421802.78255771.41209705.13295136.81275712.75267087.03331757.97327180.38307667.28364424.97120415.56318229.63
Error2886.811789.751673.861828.801899.481903.641938.981959.866596.261969.021605.321936.65
Sn(ppm)27.3048.05< LOD41.8142.6637.7636.8638.84< LOD< LOD60.3138.84
Error6.047.94N/A6.878.267.847.417.29N/AN/A8.467.05
Sr(ppm)10.448.39135.8069.53537.55518.69772.01290.42205.05245.61480.87593.59
Error1.351.283.072.537.427.179.025.144.454.856.837.36
Te(ppm)113.43228.84< LOD181.49219.32206.21212.17198.27191.17133.27282.07184.69
Error20.9127.53N/A23.6328.8727.4826.0825.4116.0812.4129.1624.46
Ti(ppm)292.67939.77956.353493.133474.302658.545768.742456.27267.943523.48980.344765.03
Error39.7643.5163.9568.2386.6588.83122.5671.1768.2084.6460.9083.48
V(ppm)35.2477.2567.2147.14155.08129.16238.74127.60145.09144.0451.38200.81
Error13.8714.4927.9619.2625.8227.4836.5222.3038.4136.1716.0624.25
Zn(ppm)30.0455.4517.3333.5236.2432.8978.0617.39135.1571.0630.3329.87
Error6.167.335.476.087.246.979.116.0210.338.206.896.48
Zr(ppm)667.30490.5983.36161.1764.6469.36104.97103.8994.31190.7059.4789.24
Error7.556.532.993.974.334.275.283.993.794.384.034.44

3 Results

The mean relative reflectance (hereafter referred to as “reflectance”) of all measured weathered and fresh surface spectral curves are displayed in Figs. 38a. For each sample, spectra of fresh and weathered surfaces, as well as the spectrum corresponding to the difference between them are plotted jointly. This spectral difference permits the detection of changes related to the slope and brightness in the spectra. Moreover, as a complementary tool, the continuum removal technique was used for detecting subtle differences among the spectra [34, 35]. Continuum removal is commonly normalized in the range of spectral reflectance from 0 to 1, which is used in spectroscopy for enhancement of the intensity of the characteristic peaks and absorption features [34, 36]. Therefore, we used continuum removal to normalize the spectral data by reducing the variation in individual spectra (Figures 38b), with the goal of overcoming differing backgrounds and interference from spectrally overlapping minerals.

3.1 Spectral Features of Rhyolite

The weathered surfaces of all rhyolitic lithologies show an overall increase in reflectance and a steep profile of the reflectance curve (Figure 3a). The diagnostic absorption features of the weathered surfaces show similar spectral behavior as the fresh surfaces, differing only slightly owing to their lower reflectance values. The spectral curve for rhyolite tends to be flat until a broad increase in absorption is visible around 1000 nm (Figure 3a). This absorption is attributed to the effect of ferrous iron and electronic transitions in discrete ions in crystal field absorption (CFA) [37, 38]. Weak absorption is visible at 480 nm, 1900 nm, and 2200 nm. Absorption by Fe ions at 480 nm produce a broad and strong peak in the absorption spectrum. The difference between fresh and weathered surfaces is the result of a weak absorption of Fe ions at 870 nm on the fresh surface. A maximum in absorption at about 1900 nm is probably due to absorption by water. Weak absorption is observable at 2200 nm, which could be caused by sericitization of plagioclase, most likely related to the presence of hydroxyl groups such as Mg–OH.

Figure 3 Weathered and fresh surface spectral curves of rhyolite (a: spectral curve; b: spectral curve with continuum removed).
Figure 3

Weathered and fresh surface spectral curves of rhyolite (a: spectral curve; b: spectral curve with continuum removed).

The differences between fresh and weathered surfaces are primarily related to spectra brightness, and are not affected by the intensity of the absorption features. In fact, the continuum-removed spectral curves have similar characteristics. However, there is an exception in the visible region, where the fresh surfaces show a more pronounced slope indicating differences in iron oxide content between the two surfaces (Figure 3b). According to XRF analysis, the Fe contents of the weathered and fresh surfaces are 13460.43±202.85 ppm and 10699.34±180.42 ppm, respectively. It is observed that the Fe content of the weathered surfaces is 1.26 times higher than that of the fresh surface.

3.2 Spectral Features of Andesite

The andesite spectral reflectance curves exhibit significant differences in spectra brightness (the whole reflectance value is much lower than other rocks) as well as in the intensity of the absorption features between fresh and weathered surfaces. It can be observed that the weathered surface reflectivity is greater than the fresh surface reflectivity from 350 nm to 1400 nm. In contrast, the reflectivity of the fresh surfaces is larger than that of the weathered surfaces (Figure 4a). The envelope diagram shows that the surface spectral absorption and change of relative reflectance values varies significantly between weathered and fresh surfaces between 350 nm and 2500 nm (Figure 4b).

Figure 4 Weathered and fresh surface spectral curves of andesite (a: spectral curve; b: the continuum removed spectral curve).
Figure 4

Weathered and fresh surface spectral curves of andesite (a: spectral curve; b: the continuum removed spectral curve).

In the visible light bands, the significant absorption spectrum of the fresh surfaces exhibits an absorption edge at 530 nm, a weak absorption valley near 645 nm, and a well-defined band centered near 880 nm. The wavelength position of the chief ferric iron absorption band (near 880 nm) and the shape of the absorption edge and shoulder at 530 nm and 645 nm are most consistent with lepidocrocite and goethite [3941].

In the short-wave infrared bands, there is a broad absorption shoulder from 1000 nm to 1950 nm with two strong absorption valleys around 1450 nm and 1940 nm. The absorption bands found near 2210 nm, 2350 nm, and 2450 nm are combinations of a ligand-OH bend and an OH stretch. Usually only two of these three bands appear. The bands appear near 2200 nm and 2300 nm when the aluminum is the bonding ligand (e.g., in dioctahedral clays). Meanwhile, Mg–OH molecules (e.g., in trioctahedral clays) give rise to absorption bands near 2300 nm and 2400 nm [42].

3.3 Spectral Features of Granite

Figure 5a shows no significant difference in spectral absorption features between weathered and fresh granite surfaces other than the slightly varying intensity of the absorption features. The spectrally characterized exterior surface was covered with a veneer of pink–brown oxidation products through which the underlying mineralogy was visible. The fresh surface appeared to be free of any alteration and the constituent minerals were readily apparent.

Figure 5 Weathered and fresh surface spectral curves of granite (a: spectral curve; b: spectral curve with continuum removed).
Figure 5

Weathered and fresh surface spectral curves of granite (a: spectral curve; b: spectral curve with continuum removed).

The granite sample differs from the others by a lower abundance of iron-bearing minerals. According to XRF analysis, the Fe content of the weathered and fresh surfaces is 34060.01±433.02 ppm and 37310.42±467.42 ppm, respectively. This is also revealed in the absence of detectable ferric iron absorption features in the spectrum of the weathered surface (Figure 5). The broad reflectance decline towards shorter wavelengths in the exterior spectrum is probably attributable to a series of charge transfers involving in a number of atomic species [42]. The inflection at 590 nm may be due to a very small amount of ferric iron, which might have derived from weathering of the biotite. While quartz is spectrally featureless, feldspar often shows a broad and weak absorption band near 1250 nm which is not evident in the weathered surfaces sample spectrum [43]. Asymmetric hydroxyl absorptions are present around at 1400 nm and 1900 nm in both spectra. Their shape is due to multiple hydrated mineral species at a range of 2205 nm, 2350 nm, and 2450 nm.

3.4 Spectral Features of Tuffaceous Sandstone

The overall shape of the reflectance curves indicates that spectral differences for this rock type were found only in the intensity of the absorption features, which are more pronounced in the weathered spectrum than in the fresh one. This could be attributed to a higher content of the iron oxides (as a result of the weathering process) and/or hydrates on the weathered surfaces (hematite or goethite) (Figure 6).

Figure 6 Weathered and fresh surface spectral curves of tuffaceous sandstone (a: spectral curve; b: spectral curve with continuum removed).
Figure 6

Weathered and fresh surface spectral curves of tuffaceous sandstone (a: spectral curve; b: spectral curve with continuum removed).

The absorption spectra of iron ions at 460 nm are in the wavelength range of visible light near the infrared region (Figure 6b). Water molecules and ferric iron together result in the formation of a new compound visible from 680 nm to 1300 nm, which is the possible reason for the slight right shift of the absorption valley. Spectral bandwidth is narrow and strong at 1400–1900 nm near a strong absorption. However, as a result of the effect of water, the two spectral bands do not have significance for mineral recognition. In the 2000–2500 nm region, the reflectance change is lower and the two spectra tend to be similar to each other. Nevertheless, the short-wave bands at 2200 nm and 2350 nm have different spectral absorption characteristics, caused by hydroxyl, CO32, and composite spectra among which the carbonate ion absorption bands are relatively stronger [6, 44]. Meanwhile, the strong absorption at 2350 nm in CO32 is due to a higher content of sheet silicates on the fresh rock surfaces [16, 42]

3.5 Spectral Features of Basalt

The reflectance spectrum of weathered basalt surfaces is characterized by ferric iron absorption bands visible as an abrupt rise in reflectance near 530 nm, and two broad absorption bands near 730 nm and 920 nm. At longer wavelengths, an absorption feature is present near 1090 nm and there is an additional broad, intricate absorption feature between 1900 nm and 2500 nm. The reflectance shows maxima at 2250 nm and 2350 nm, which are followed by a reflectance decrease toward longer wavelengths with a shoulder at 2200 nm, and minima at 1910 nm, 1990 nm, and 2450 nm (Figure 7).

Figure 7 Weathered and fresh surface spectral curves of basalt (a: spectral curve; b: spectral curve with continuum removed).
Figure 7

Weathered and fresh surface spectral curves of basalt (a: spectral curve; b: spectral curve with continuum removed).

The ferric iron absorption bands are approximately focused at 480 nm, 730 nm, 920 nm, and 1090 nm. However, their position cannot be precisely determined. The broadness of this feature is consistent with multiple, hydrated ferric iron phases such as goethite, feroxyhyte, and ferrihydrite [44, 45]. Strong ferric iron absorption bands are expected for weathering products of iron-rich materials such as basalt. The weathered surface spectrum is not as strongly affected by ferric iron absorption bands as the fresh spectrum. Relatively, a broad band is centered near 1090 nm, which shows the presence of ferrous rather than ferric iron in the fresh spectrum [37, 46].

At longer wavelengths, the fresh and weathered surface spectra exhibit similar spectral features. All of these bands can be due to various hydrated mineral phases. The bands are relatively more intense in the weathered surface spectrum. The 1900 nm band is broad and shows a complex shape, which is consistent with multiple hydrated species and/or water adsorbed onto the surface of samples from a number of sites. Structural OH is probably present in only one or two sites, which may contribute to the sharpness of the 1400 nm feature. There is also a contribution of adsorbed water to this band; however, the relative abundances of the two types of water (bound water and structural water) cannot be determined. The 2200–2500 nm regions show strong absorption with discrete absorption features at 2200 nm, 2250 nm, 2350 nm, and 2450 nm, which are characteristics for Al–OH and Mg–OH. Both Al and Mg are present in basalts, and consequently this type of spectral behavior is expected in the weathered surface spectra. Although precise spectral matching is not feasible, mineral phases such as gibbsite, chlorite, and serpentine provide the best spectral matches to these bands. Overlaps and interferences of absorption bands occur in multicomponent mineral mixtures [47], which makes it difficult to determine the relative contributions of each phase.

3.6 Spectral Features of Diorite

The diorite spectral reflectance curves show that the reflectivity of fresh surfaces is greater than that of weathered surface from 350 nm to 990 nm. In contrast, the reflectivity of fresh surfaces is larger than that of weathered surface from 990 nm to 2500 nm (Figure 8a).

Figure 8 Weathered and fresh surface spectral curves of diorite (a: spectral curve; b: spectral curve with continuum removed).
Figure 8

Weathered and fresh surface spectral curves of diorite (a: spectral curve; b: spectral curve with continuum removed).

As observed from the continuum removal diagram (Figure 8b), there is an absorption edge near 390 nm, and the weathered surfaces have lower reflectance values than the fresh surfaces. A broad and symmetric band that appears in the 1000 nm region could be caused by Fe2+ –Fe3+crystal field transitions on the weathered and fresh surfaces [48]. Broad hydroxyl bands are present near 1400 nm and 1900 nm. These bands are considerably stronger and broader than those of biotite and cordierite, suggesting the possibility of an optically significant weathered zone. A local reflectance peak is present at 2200 nm, followed by a decrease in reflectance with absorption features superimposed at 2250 nm, 2300 nm, 2380 nm, and 2450 nm. Correlation of these bands with specific minerals is hampered by the observation that different samples of the same mineral can have absorption bands at different wavelength positions and of different relative strengths [36, 37, 4951]. The absorption bands in the sample spectrum appear to be caused by weathering products that consist of both Al–OH and Mg–OH absorption bands. This is supported by the observation that the lattice -OH bands are not very distinct and a certain degree of disorder must be present in weathered materials.

As the differences between fresh and weathered surfaces are mainly related to spectra brightness and do not affect the intensity of the absorption features, the continuum removal curves for both weathered and fresh surfaces are similar. The fresh surfaces present a more pronounced slope, indicating the differences in iron oxide/hydroxide content between both surfaces.

4 Discussion

Depending on environmental agents (climate and topography) and lithology (mineralogical composition of the original rock), most exposed rock surfaces undergo a certain degree of alteration and/or weathering. In the West Junggar region, there are significant temperature variations and strong wind gradients. Most of the exposed rock formations in this region are covered with weathered layers of different degrees. These weathering processes alter the shape, overall reflectance values, and certain absorption characteristics of different rock types in arid and semi-arid environments.

Weathering processes result in the formation of new materials that differ both chemically and mechanically from the original parent rock [13, 42]. New materials formed by physical weathering will normally have a mineral composition similar to the parent rock. Some of the weathering products are less massive and less indurated, possessing much lower mechanical strength [16, 42]. The present study reveals that, for better spectral characterization, recognition and discrimination of lithological units are necessary, which is possible from remotely sensed data collected in the field. The spectral regions where fresh and weathered surfaces show spectral fluctuations can be used to better characterize and discriminate lithological units [48, 54]. This study demonstrates the spectral differences between weathered and fresh surfaces caused by mineralogical changes induced by weathering. The spectral characteristics and the continuum-removed spectral curves reveal that the features of the fresh and weathered surface differs to varying degrees. At high spectral resolution, differences are minimal for rhyolite, granite, and tuffaceous sandstone, but large for andesite, basalt, and diorite.

Spectra of fresh and weathered surfaces were surveyed and are discussed in terms of their differences in compositional and textural characteristics [13, 55, 56]. The most important differences observed between fresh and weathered surface spectra are related to (1) spectrum brightness, (2) presence and intensity of characteristic absorption features, and (3) the differences of slope in the 350 – 2500 nm region [13, 22, 57]. The present study shows that spectral changes occur mainly in the 350–1000 nm and 1900–2500 nm regions.

In arid and semi-arid environments, the most common alteration phenomena are the formation of clays and iron oxide/hydroxides. The weathered sample spectra show numerous features typical of secondary weathering minerals, which are inconsistent with the underlying, more primary mineralogy. Even when the visible depth of weathering is less than a few hundred micrometers, secondary weathering minerals (mainly hematite and/or goethite) dominate the reflectance spectrum. In most of the iron-rich samples, the visible-wavelength regions of the spectra for weathered samples are dominated by a change in the absorption features and slope of the spectra in the ferric iron absorption bands. Previous studies have also documented absorption features of rocks in the visible part of the reflectance spectra caused by the presence of different iron minerals [37, 58], as well as various mixtures of ferric oxides/hydroxides and other minerals [44, 5961]. Different ferric oxides are identified by their unique spectral properties such as an absorption edge near 530 nm (low reflectance below this wavelength), an absorption band or shoulder near 640 nm, and a strong absorption band between 860 and 980 nm. These features are attributable to crystal field transitions and the wavelength position of the latter band, which is the characteristic feature of a particular ferric iron-bearing mineral present rocks [17, 44, 58, 62, 63]. Studies of various ferric oxide mineral mixtures indicate that ferric minerals dominate the spectrum at shorter wavelengths of less than 1000 nm, while at longer wavelengths other minerals dominate the spectrum [42, 44, 60]. There appears to be a strong band in the 1000 nm region, which indicates that ferrous rather than ferric iron is present in the fresh spectrum of rocks such as tuffaceous sandstones (XM78) and basalts (XM04). This phenomenon may also be attributed to a secondary alteration of iron-rich minerals, similar to what has been observed in deeply weathered volcano-sedimentary rocks [16].

In the short-wave infrared region (1000–2500 nm), reflectance spectra of clay minerals show a number of features that have also been documented in previous studies [3, 37, 51]. It is obvious that spectral features at 1000–2500 nm are due to O–H vibrations, features at 2200–2500 nm are solely due to hydroxyl groups, and the feature at 1900 nm is caused by thewater content of the rocks. When water is incorporated into the crystal lattice as-structural water bonded to Al–OH or Mg–OH, absorption bands are predictable to occur in the range of 1900 nm and 2500 nm [46, 47].

The strongest Al–OH bands appear near 2200 nm, while Mg–OH bands are found near 2300 nm and 2350 nm. Mg–OH bands are most evident in many of the magnesium-rich samples (basalt and diorite), while most of the aluminum- and magnesium-deficient samples (granite) are devoid of resolvable Al–OH and Mg–OH bands. The appearance of Al–OH and Mg–OH absorption bands in exterior surface spectra is also correlated with the presence of appreciable concentrations of Al and Mg in the underlying lithology. Difficulties exist in the determination and discrimination of a particular weathering product because of the lack of unambiguous lattice–OH absorption features and overlaps by adjacent bands.

All spectra exhibit a number of characteristic absorption bands, with bound water resulting in absorption bands mainly near 1400 nm and 1900 nm, and structural water resulting in absorption bands near 1400 nm, 2200 nm, 2300 nm, and 2400 nm. The widths, intensities, and shapes of the 1400 nm and 1900 nm bands are related to the geometries of the sites occupied by water. Narrow and symmetric bands are characteristic of water molecules situated in one or a few well–ordered sites, while broad or complex bands correspond to disordered or multiple sites [36, 38]. In this discussion, we have not considered features located around 1400 nm and 1900 nm, as they are caused by atmospheric water absorbed by the samples.

The spectral differences presented in this paper include spectral slopes, the appearance/disappearance of absorption bands, shifts in absorption band minimum wavelength positions, and changes in band shape. The obtained results provide valuable information for the geological analysis of airborne long-wave hyperspectral imagery, as in the case of SEBASS and EO–1 Hyperion. Much of the analysis of hyperspectral data for geological applications is based on the detection and identification of absorption features and their intensities in a narrow band range [6467]. However, reflectance (brightness) and spectra slope differences can also be assessed by analysis of broad spectral bands as in the case of Landsat Thematic Mapper. For this reason, spectral differences between fresh and weathered surfaces were also examined for the ASTER and Landsat–TM bands by filtering the spectra for the relative spectral response of that sensor [6870].

In future studies, hyperspectral data from the spaceborne Earth Observing–1 (EO–1) Hyperion satellite may provide a more viable and reliable alternative to existing data. The high-resolution spectral and spatial data from Hyperion yield repeatable images of high quality, which may allow accurate mineral mapping at a large scale. Excluding those hyperspectral bands that are uncalibrated or disturbed by vapor, there are 196 unique channels (8–57 for the visible and near-infrared (VNIR) bands and 77–224 for the short-wave infrared (SWIR) bands) available from Hyperion [7173]. Thus, the Hyperion hyperspectral bands covering 426–926 nm (bands 8–57) and 1942–2385 nm (bands 179–223) match the wavelengths of iron oxides and hydroxides within the domains of 400–900 nm and 1900–2400 nm. In further studies, we will use high-resolution spectral and spatial data from Hyperion.

On the other hand, multispectral remote sensing images have the potential for mineral mapping using broad bandwidths (low spectral resolution). For example, ASTER has 14 spectral bands covering visible (VIS) and near-infrared (NIR) (bands 1–3), SWIR (bands 4–9), and thermal infrared TIR spectra (bands 10–14). Thus, the wavelengths of iron oxides and hydroxides derived in the present study from VIS and NIR (400–900 nm) and SWIR (2000–2450 nm) show a close correlation with band 1 (520–600 nm), band 2 (630–690 nm), band 3 (780–860 nm), and SWIR band 5 (2145–2185 nm), band 6 (2185–2225 nm), band 7 (2235–2285 nm), band 8 (2295–2365 nm) and band 9 (2360–2430 nm) of ASTER, respectively [7479]. Especially the five SWIR bands correlate well with the wavelengths of hydroxides. Similarly, Landsat 8 OLI has nine spectral bands covering the VIS and NIR (bands 1–5), the SWIR (bands 6, 7, and 9), and panchromatic band (band 8) (70, 80, 81). Therefore, in further studies, wavelengths of the blue band 1 (433–453 nm), band 2 (450–515 nm), band 3 (525–600 nm), band 4 (630–680 nm), band 5 (845–885 nm), and SWIR band 7 (2100–2300 nm) of Landsat 8 OLI can be utilized. The bands of these and other satellites closely match the wavelengths of iron oxides and hydroxides; however, their use has had limited success because of their low spatial and spectral resolutions.

With continuing development of hyperspectral and multispectral remote sensing techniques, mineral recognition has improved from approximate to detailed, qualitative to quantitative, and indirect to direct. In order to understand the behavior of a mineral spectrum, we must minimize or eliminate external influences on spectral information such as mineral particle size, geometrical variations of optical sensor orientation, state of the mineral surface (desert varnish, overgrowths, etc.), and weathering effects (physical and chemical). In the near future, hyperspectral remote sensing technology can be used to better perform mineral recognition, differentiation of similar minerals, and analysis of mineral content and chemical composition.

Surface weathering produces coatings by adherence of wind-borne dust, cemented coatings, and attachment of particles on rock surfaces, which are importance keys for the interpretation of reflectance spectra.

5 Conclusions

Surfaces weathering of rocks in which mineral materials may be similar to or quite different from the minerals in the underlying parent rock completely control the reflectance spectra of the terrain. Analysis of these coatings is important for the application of reflectance spectra to geologic mapping. This study contributes ASD field spectral radiometer data, X-ray fluorescence analyses, and petrographic observations of weathered and fresh rock samples (rhyolite, andesite, granite, tuffaceous sandstone, basalt, and diorite) in Xiemisitai, Western Junggar region, Xinjiang. In arid and semi-arid environments, materials formed by weathering processes differ both chemically and mechanically from the original parent rock, and induce changes in the spectral features of fresh and weathered rock surfaces. These alterations include variations in overall spectral slope, appearance/disappearance of absorption bands, shifts absorption band at different wavelength positions, and changes in band shape. This paper presents spectral characteristics and continuum-removed spectral curves that indicate small spectral differences between weathered and fresh surfaces for rhyolite, granite, and tuffaceous sandstone, but large differences for andesite, basalt, and diorite. Spectral characteristics in the 350–1000 nm region are attributed to alteration of preexisting iron oxide minerals by atmospheric processes or secondary alteration of iron-rich minerals. In the short-wave infrared region, spectral features between 1000–2500 nm are due to O–H vibrations, with features at 2200–2500 nm solely caused by hydroxyl groups. The strongest Al–OH bands appear near 2200 nm, while Mg–OH bands are found near 2300 nm and 2350 nm. The presented spectroscopic results for fresh and weathered rock samples can be applied to better characterize and discriminate lithological units and to mineral mapping using hyperspectral and multispectral remote sensing images.

Acknowledgement

This work was supported by the National Natural Science Foundation of China (Grant No. 41402296), the National Natural Science Union Foundation of China (Grant No. U1503291), and under the Science and Technology Plan Major Projects of the Xinjiang Uygur Autonomous Region (Grant No. 201330121-2).

References

[1] Feng J., Rivard B., Rogge D., and Sánchez-Azofeifa A., The long-wave infrared (3–14 μm) spectral properties of rock encrusting lichens based on laboratory spectra and airborne SEBASS imagery. (2013), Remote Sensing of Environment 131(173-181)10.1016/j.rse.2012.12.018Search in Google Scholar

[2] Michalski J.R., Kraft M.D., Sharp T.G., and Christensen P.R.,Effects of chemical weathering on infrared spectra of Columbia River Basalt and spectral interpretations of martian alteration. (2006), Earth and Planetary Science Letters 248(3–4), 822-82910.1016/j.epsl.2006.06.034Search in Google Scholar

[3] Hunt G.R., Spectroscopic properties of rocks and minerals.(1982), Crc Handbook of Physical Properties of Rocks 8(4), 658Search in Google Scholar

[4] Hueni A., Nieke J., Schopfer J., Kneubühler M., and Itten K.I., The spectral database SPECCHIO for improved long-term usability and datasharing. (2009), Computers & Geosciences 35(3), 557-56510.1016/j.cageo.2008.03.015Search in Google Scholar

[5] Milton E.J., Schaepman M.E., Anderson K., Kneubühler M., and Fox N., Progress in field spectroscopy. (2009), Remote Sensing of Environment 113, Supplement 1 (S92-S109)10.1109/IGARSS.2006.509Search in Google Scholar

[6] Clark R.N., Swayze G.A., Livo K.E., Kokaly R.F., Sutley S.J.,Dalton J. B., McDougal R.R., and Gent C.A., Imaging spectroscopy: Earth and planetary remote sensing with the USGS Tetracorder and expert systems. (2003), Journal of Geophysical Research: Planets 108(E12), 513110.1029/2002JE001847Search in Google Scholar

[7] Christensen P.R., Bandfield J.L., Hamilton V.E., Howard D.A.,Lane M.D., Piatek J.L., Ruff S.W., and Stefanov W.L., A thermal emission spectrallibrary of rock-forming minerals. (2000), Journal of GeophysicalResearch-Planets 105(E4), 9735-973910.1029/1998JE000624Search in Google Scholar

[8] Baldridge A.M., Hook S.J., Grove C.I., and Rivera G., The ASTER-spectral library version 2.0. (2009), Remote Sensing of Environment 113(4), 711-71510.1016/j.rse.2008.11.007Search in Google Scholar

[9] Burrows J.P., Weber M., Buchwitz M., Rozanov V., Ladstatter-Weissenmayer A., Richter A., DeBeek R., Hoogen R., Bramstedt K., Eichmann K.U., and Eisinger M., The global ozone monitoring experiment (GOME):Mission concept and first scientific results. (1999), Journal of the Atmospheric Sciences 56(2), 151-17510.1175/1520-0469(1999)056<0151:TGOMEG>2.0.CO;2Search in Google Scholar

[10] World Agroforestry C., and Isric, ICRAF-ISRIC Soil VNIR Spectral-Library, ICRAF, ISRIC (2010).Search in Google Scholar

[11] Rivard B., Arvidson R.E., Duncan I.J., Sultan M., and Elkaliouby B., Varnish, sediment, and rock controls on spectral reflectance of outcrops inarid regions. (1992), Geology 20(4), 295-29810.1130/0091-7613(1992)020<0295:VSARCO>2.3.CO;2Search in Google Scholar

[12] Rowan L.C., Simpson C.J., and Mars J.C., Hyperspectral analysis of the ultramafic complex and adjacent lithologies at Mordor, NT, Australia.(2004), Remote Sensing of Environment 91(3–4), 419-43110.1016/j.rse.2004.04.007Search in Google Scholar

[13] Younis M.T., Gilabert M.A., Melia J., and Bastida J., Weathering process effects on spectral reflectance of rocks in a semi-arid environment.(1997), International Journal of Remote Sensing 18(16), 3361-337710.1080/014311697216928Search in Google Scholar

[14] Lyon R., and Honey F., Spectral Signature Extraction From Airborne Imagery Using The Geoscan MkII Advanced Airborne Scanner In The Leonora, Western Australia Gold District. (1989), Geoscience and Remote Sensing Symposium, 1989. IGARSS’89. 12th Canadian Symposium on Remote Sensing., 1989 International 2925-293010.1109/IGARSS.1989.575955Search in Google Scholar

[15] Lyon R.J.P., and Honey F.R., Direct mineral identification with-Geoscan Mk II Advanced Multispectral Scanner. (1990), Imaging Spectroscopy of the Terrestrial Environment 50-6110.1117/12.21335Search in Google Scholar

[16] Metelka V., Baratoux L., Jessell M.W., and Naba S., Visible and infrared properties of unaltered to weathered rocks from Precambriangranite-greenstone terrains of the West African Craton. (2015), Journal of African Earth Sciences10.1016/j.jafrearsci.2015.10.003Search in Google Scholar

[17] Singer R.B., Spectral evidence for the mineralogy of high-albedosoils and dust on Mars. (1982), Journal of Geophysical Research SolidEarth 87(B12), 10159–1016810.1029/JB087iB12p10159Search in Google Scholar

[18] Rogge D., Rivard B., Segl K., Grant B., and Feng J., Mapping of NiCu–PGE ore hosting ultramafic rocks using airborne and simulated EnMAPhyperspectral imagery, Nunavik, Canada. (2014), Remote Sensing of Environment 152 (302-317)10.1016/j.rse.2014.06.024Search in Google Scholar

[19] Zhang J., Rivard B., and Sánchez-Azofeifa A., Spectralunmixing of normalized reflectance data for the deconvolution of lichen and rockmixtures. (2005), Remote Sensing of Environment 95(1), 57-6610.1016/j.rse.2004.11.019Search in Google Scholar

[20] Kuhlman K., Venkat P., La Duc M., Kuhlman G., and Mckay C.,Evidence of a microbial community associated with rock varnish at Yungay, AtacamaDesert, Chile. (2008), Journal of Geophysical Research Biogeosciences 113(G4), 285-29510.1029/2007JG000677Search in Google Scholar

[21] Kuhlman K.R., Allenbach L.B., Ball C.L., Fusco W.G., La DucM.T., Kuhlman G.M., Anderson R.C., Stuecker T., Erickson I.K., Benardini J., andCrawford R.L., Enumeration, isolation, and characterization of ultraviolet (UV-C)resistant bacteria from rock varnish in the Whipple Mountains, California.(2005), Icarus 174(2), 585-59510.1016/j.icarus.2004.11.022Search in Google Scholar

[22] Wayne D.M., Diaz T.A., Fairhurst R.J., Orndorff R.L., and Pete D.V., Direct major- and trace-element analyses of rock varnish by high resolution laser ablation inductively-coupled plasma mass spectrometry (LA-ICPMS). (2006), Applied Geochemistry 21(8), 1410-143110.1016/j.apgeochem.2006.04.005Search in Google Scholar

[23] Lyon R.J.P., The Effects of Weathering and Other Varnish-like Coatings on High Spectral Resolution Airborne Imagery (Part one). (1996),Environmental Remote SensingSearch in Google Scholar

[24] Lyon R.J.P., Effects Of Weathering, Desert-Varnish, Etc. On Spectral Signatures Of Mafic, Ultramafic And Felsic Rocks, Leonora, West Australia. (1990), Digest–International Geoscience and Remote Sensing Symposium (IGARSS) 1719-172210.1109/IGARSS.1990.688846Search in Google Scholar

[25] Wang S., Zhou K., Zhang N., and Wang J., Spectral data analysis of rock and mineral in Hatu Western Junggar Region, Xinjiang. (2014), SPIE Asia Pacific Remote Sensing10.1117/12.2068697Search in Google Scholar

[26] Mao W.Y., Zhang X., Yang Z.H., Sayran, Chen P.X., and Shen Y.P.,Winter Snowmelt Flood in the Northwest of Junggar Basin First Monitored bySatellite Remote Sensing. (2010), Journal of Glaciology & Geocryology32(1), 211-214Search in Google Scholar

[27] Zhang W.S., Huang J.G., Zari M., and Gao H.Z., Remote sensingore-hunting information of end genetic gold deposits in eastern Junggar region, Xinjiang. (2003), Geological Science & Technology Information 22(2),77-81Search in Google Scholar

[28] Liu L., Zhou J., Jiang D., Zhuang D., Mansaray L.R., and Zhang B., Targeting Mineral Resources with Remote Sensing and Field Data in the Xiemisitai Area, West Junggar, Xinjiang, China. (2013), Remote Sensing10.3390/rs5073156Search in Google Scholar

[29] Shen P., Shen Y., Li X.H., Pan H., Zhu H., Meng L., and Dai H., Northwestern Junggar Basin, Xiemisitai Mountains, China: A geochemical and geochronological approach. (2012), Lithos s 140–141(5), 103–11810.1016/j.lithos.2012.02.004Search in Google Scholar

[30] Chen B., and Arakawa Y., Elemental and Nd-Sr isotopic geochemistry of granitoids from the West Junggar foldbelt (NW China), with implications for Phanerozoic continental growth. (2005), Geochimica EtCosmochimica Acta 69(5), 1307-132010.1016/j.gca.2004.09.019Search in Google Scholar

[31] Xiao W., Han C., Chao Y., Min S., Lin S., Chen H., Li Z., Li J., and Shu S., Middle Cambrian to Permian subduction-related accretionary orogenesis of Northern Xinjiang, NW China: Implications for the tectonic evolution of central Asia. (2008), Journal of Asian Earth Sciences 32(2–4), 102-11710.1016/j.jseaes.2007.10.008Search in Google Scholar

[32] Shen P., Shen Y., Liu T., Meng L., Dai H., and Yang Y., Geochemical signature of porphyries in the Baogutu porphyry copper belt, western Junggar, NW China. (2009), Gondwana Research 16(2), 227-24210.1016/j.gr.2009.04.004Search in Google Scholar

[33] Zhang N.N., and Zhou K.F., Identification of hydrothermal alteration zones of the Baogutu porphyry copper deposits in northwest China usingASTER data. (2017), Journal of Application Remote Sesening 11(1), doi:10.1117/1.JRS.11.015016.10.1117/1.JRS.11.015016Search in Google Scholar

[34] Clark R.N., King T.V.V., Klejwa M., Swayze G.A., and Vergo N.,High spectral resolution reflectance spectroscopy of minerals. (1990), Journal of Geophysical Research: Solid Earth 95(B8), 12653-1268010.1029/JB095iB08p12653Search in Google Scholar

[35] Rodger A., Laukamp C., Haest M., and Cudahy T., A simple-quadratic method of absorption feature wavelength estimation in continuum removedspectra. (2012), Remote Sensing of Environment 118(0), 273-28310.1016/j.rse.2011.11.025Search in Google Scholar

[36] Clark R.N., and Roush T.L., Reflectance spectroscopy:Quantitative analysis techniques for remote sensing applications. (1984), Journal of Geophysical Research Atmospheres 89(B7), 6329-634010.1029/JB089iB07p06329Search in Google Scholar

[37] Hunt G.R., Near-infrared (1.3 2.4) μm spectra of alteration minerals–Potential for use in remote sensing. (1979), Geophysics 44(12), 197410.1190/1.1440951Search in Google Scholar

[38] Hunt G.R., and Ashley R.P., Spectra of altered rocks in thevisble and near-infrared. (1979), Economic Geology 74(7), 1613-162910.2113/gsecongeo.74.7.1613Search in Google Scholar

[39] Velasco F., Alvaro A., Suarez S., Herrero J.-M., and Yusta I., Mapping Fe-bearing hydrated sulphate minerals with short wave infrared (SWIR)spectral analysis at San Miguel mine environment, Iberian Pyrite Belt (SW Spain).(2005), Journal of Geochemical Exploration 87(2), 45-7210.1016/j.gexplo.2005.07.002Search in Google Scholar

[40] Chevrier V., Roy R., Le Mouélic S., Borschneck D., MathéP.E., and Rochette P., Spectral characterization of weathering products of elemental iron in a Martian atmosphere: Implications for Mars hyperspectralstudies. (2006), Planetary and Space Science 54(11), 1034-104510.1016/j.pss.2005.12.019Search in Google Scholar

[41] Sellitto V.M., Fernandes R.B.A., Barrón V., and Colombo C.,Comparing two different spectroscopic techniques for the characterization of soiliron oxides: Diffuse versus bi-directional reflectance. (2009), Geoderma 149(1–2), 2-910.1016/j.geoderma.2008.11.020Search in Google Scholar

[42] Cloutis E.A., Weathered and unweathered surface spectra of rocks from cold deserts: Identification of weathering processes and remote sensing implications. (1992), Geologiska Föreningen I StockholmFörhandlingar 114(2), 181-19110.1080/11035899209453883Search in Google Scholar

[43] Ding C., Liu X., Liu W., Liu M., and Li Y., Mafic–ultramafic and quartz-rich rock indices deduced from ASTER thermal infrared data using alinear approximation to the Planck function. (2014), Ore Geology Reviews60(0), 161-17310.1016/j.oregeorev.2014.01.005Search in Google Scholar

[44] Sherman D.M., Burns R.G., and Burns V.M., Spectral characteristics of the iron oxides with application to the Martian bright regionmineralogy. (1982), Journal of Geophysical Research Solid Earth 87(B12),10169–1018010.1029/JB087iB12p10169Search in Google Scholar

[45] Thomas S., and Sherwood P.M.A., Valence band spectra of aluminumoxides, hydroxides, and oxyhydroxides interpreted by X.alpha. calculations.(1992), Analytical Chemistry 64(21), 2488-249510.1021/ac00045a006Search in Google Scholar

[46] Cloutis E.A., Gaffey M.J., Lambert J., and Smith D.G.W., "The-quality of geological information derivable from high resolution reflectance spectra–Results for mafic silicates," (1987), 10th Canadian Symposium on Remote SensingSearch in Google Scholar

[47] Cloutis E.A., Gaffey M.J., Smith D.G.W., and Lambert R.S.J., Reflectance spectra of mafic silicate-opaque assemblages with applications tometeorite spectra. (1990), Icarus 84(2), 315-33310.1016/0019-1035(90)90041-7Search in Google Scholar

[48] Johnson J.R., Grundy W.M., and Shepard M.K., Visible/near-infrared spectrogoniometric observations and modeling of dust-coatedrocks. (2004), Icarus 171(2), 546-55610.1016/j.icarus.2004.05.013Search in Google Scholar

[49] Goetz A.F., Vane G., Solomon J.E., and Rock B.N., Imaging spectrometry for Earth remote sensing. (1985), Science 228(4704),1147-115310.1126/science.228.4704.1147Search in Google Scholar

[50] Roush T.L., Singer R.B., and Mccord T.B., “Reflectance Spectraof Selected Mafic Silicates From .6 to 4.6 microns,” (1998), Lunar and Planetary Science ConferenceSearch in Google Scholar

[51] Clark R.N., “Spectroscopy of rocks and minerals, and principles of spectroscopy,” in Remote sensing for the earth sciences: Manual ofremote sensing,. (1999), pp. 3-58, John Wiley and Sons, New YorkSearch in Google Scholar

[52] Rivard B., and Arvidson R.E., Utility of imaging spectrometry for lithologic mapping in greenland (1992), Photogrammetric Engineering and Remote Sensing 58(7), 945-949Search in Google Scholar

[53] Dorn R.I., Rock Varnish. (2008), American Scientist 79(6), 246-29710.1177/030913338200600301Search in Google Scholar

[54] Feng J., Rivard B., Rogge D., and Sánchez-Azofeifa A., The long-wave infrared (3–14; μm) spectral properties of rock encrusting lichens based on laboratory spectra and airborne SEBASS imagery. (2013), Remote Sensing of Environment 131(0), 173-18110.1016/j.rse.2012.12.018Search in Google Scholar

[55] Rawlins B.G., Kemp S.J., and Milodowski A.E., Relationships between particle size distribution and VNIR reflectance spectra are weaker for soils formed from bedrock compared to transported parent materials. (2011), Geoderma 166(1), 84-9110.1016/j.geoderma.2011.07.015Search in Google Scholar

[56] Crouvi O., Ben-Dor E., Beyth M., Avigad D., and Amit R.,Quantitative mapping of arid alluvial fan surfaces using field spectrometer and hyperspectral remote sensing. (2006), Remote Sensing of Environment 104(1), 103-11710.1016/j.rse.2006.05.004Search in Google Scholar

[57] Karnieli A., Kidron G.J., Glaesser C., and Ben-Dor E., Spectral Characteristics of Cyanobacteria Soil Crust in Semiarid Environments. (1999), Remote Sensing of Environment 69(1), 67-7510.1016/S0034-4257(98)00110-2Search in Google Scholar

[58] Townsend T.E., Discrimination of iron alteration minerals invisible and near-infrared reflectance data. (1987), Journal of Geophysical Research Atmospheres 92(B2), 1441-145410.1029/JB092iB02p01441Search in Google Scholar

[59] Sherman D.M., and Waite T.D., Electronic spectra of Fe3+ oxides and oxide hydroxides in the near IR and near UV.(1985), American Mineralogist 70(11-12), 1262-126910.1007/BF00308210Search in Google Scholar

[60] Singer R.B., Near-infrared spectral reflectance of mineral mixtures: Systematic combinations of pyroxenes, olivine, and iron oxides. (1981), Journal of Geophysical Research: Solid Earth (1978–2012) 86(B9),7967–798210.1029/JB086iB09p07967Search in Google Scholar

[61] Singer R.B., Spectroscopic observation of Mars. (1985), Advances in Space Research 5(8), 59-6810.1016/0273-1177(85)90242-XSearch in Google Scholar

[62] Morris R.V., Lauer H.V., Lawson C.A., Gibson E.K., Nace G.A., and Stewart C., Spectral and other physicochemical properties of submicronpowders of hematite (α-Fe2O3), maghemite (γ-Fe2O3), magnetite (Fe3 O4), goethite (α-FeOOH), and lepidocrocite (γ-FeOOH). (1985), Journal of Geophysical Research-Solid Earth 90(B4), 3126–314410.1029/JB090iB04p03126Search in Google Scholar

[63] Sagan C., Phaneuf J.P., and Ihnat M., Total reflection spectrophotometry and thermogravimetric analysis of simulated Martian surfacematerials. (1965), Icarus 4(1), 43-6110.1016/0019-1035(65)90017-5Search in Google Scholar

[64] Magendran T., and Sanjeevi S., Hyperion image analysis and linear spectral unmixing to evaluate the grades of iron ores in parts of Noamundi, Eastern India. (2014), International Journal of Applied Earth Observation and Geoinformation 26(0), 413-42610.1016/j.jag.2013.09.004Search in Google Scholar

[65] Kodikara G.R.L., Woldai T., van Ruitenbeek F.J.A., Kuria Z., vander Meer F., Shepherd K.D., and van Hummel G.J., Hyperspectral remote sensing of evaporate minerals and associated sediments in Lake Magadi area, Kenya. (2012), International Journal of Applied Earth Observation and Geoinformation 14(1), 22-3210.1016/j.jag.2011.08.009Search in Google Scholar

[66] Magendran T., Statistical and neural network analysis of hyperspectral radiometric data to characterise hematite of Singbhum iron orebelt, India. (2014), SPIE201410.1117/12.2070388Search in Google Scholar

[67] Murphy R.J., and Monteiro S.T., Mapping the distribution offerric iron minerals on a vertical mine face using derivative analysis of hyperspectral imagery. (2013), ISPRS Journal of Photogrammetry and Remote Sensing 75(0), 29-3910.1016/j.isprsjprs.2012.09.014Search in Google Scholar

[68] Zhang X., Pazner M., and Duke N., Lithologic and mineral information extraction for gold exploration using ASTER data in the south Chocolate Mountains (California). (2007), ISPRS Journal of Photogrammetry and Remote Sensing 62(4), 271-28210.1016/j.isprsjprs.2007.04.004Search in Google Scholar

[69] Son Y.-S., Kang M.-K., and Yoon W.-J., Lithological and mineralogical survey of the Oyu Tolgoi region, Southeastern Gobi, Mongolia using ASTER reflectance and emissivity data. (2014), International Journal of Applied Earth Observation and Geoinformation 26(0), 205-21610.1016/j.jag.2013.07.004Search in Google Scholar

[70] Rajesh H.M., Mapping Proterozoic unconformity-related uranium deposits in the Rockhole area, Northern Territory, Australia using landsat ETM+.(2008), Ore Geology Reviews 33(3–4), 382-39610.1016/j.oregeorev.2007.02.003Search in Google Scholar

[71] Xu Y., Zhang Z., and Hu G., Ground object identification-based on absorption-band position using EO-1 hyperion data. (2010), Journal ofthe Indian Society of Remote Sensing 38(2), 169-17810.1007/s12524-010-0016-8Search in Google Scholar

[72] Kruse F.A., Boardman J.W., and Huntington J.F., Comparison of airborne hyperspectral data and EO-1 Hyperion for mineral mapping. (2003), Geoscience and Remote Sensing, IEEE Transactions on 41(6), 1388-140010.1109/TGRS.2003.812908Search in Google Scholar

[73] Hosseinjani Zadeh M., Tangestani M.H., Roldan F.V., and YustaI., Sub-pixel mineral mapping of a porphyry copper belt using EO-1 Hyperion data.(2014), Advances in Space Research 53(3), 440-45110.1016/j.asr.2013.11.029Search in Google Scholar

[74] Pour A.B., and Hashim M., The application of ASTER remote-sensing data to porphyry copper and epithermal gold deposits. (2012), OreGeology Reviews 44(0), 1-910.1016/j.oregeorev.2011.09.009Search in Google Scholar

[75] Moghtaderi A., Moore F., and Mohammadzadeh A., The application of advanced space-borne thermal emission and reflection (ASTER) radiometer datain the detection of alteration in the Chadormalu paleocrater, Bafq region,Central Iran. (2007), Journal of Asian Earth Sciences 30(2), 238-25210.1016/j.jseaes.2006.09.004Search in Google Scholar

[76] Mars J.C., and Rowan L.C., Spectral assessment of new ASTER SWIR surface reflectance data products for spectroscopic mapping of rocks andminerals. (2010), Remote Sensing of Environment 114(9), 2011-202510.1016/j.rse.2010.04.008Search in Google Scholar

[77] Cui J., Yan B., Wang R.S., Tian F., Zhao Y.J., Liu D.C., Yang S.M., Shen W., Regional-scale mineral mappingusing ASTER VNIR/SWIR data and validation of reflectance and mineral map productsusing airborne hyperspectral CASI/SASI data. (2014), InternationalJournal of Applied Earth Observation and Geoinformation 33(127-141)10.1016/j.jag.2014.04.014Search in Google Scholar

[78] Di Tommaso I., and Rubinstein N., Hydrothermal alteration mapping using ASTER data in the Infiernillo porphyry deposit, Argentina. (2007), Ore Geology Reviews 32(1–2), 275-29010.1016/j.oregeorev.2006.05.004Search in Google Scholar

[79] Azizi H., Tarverdi M.A., and Akbarpour A., Extraction of hydrothermal alterations from ASTER SWIR data from east Zanjan, northern Iran.(2010), Advances in Space Research 46(1), 99-10910.1016/j.asr.2010.03.014Search in Google Scholar

[80] Sadeghi B., Khalajmasoumi M., Afzal P., Moarefvand P., Yasrebi A.B., Wetherelt A., Foster P., and Ziazarifi A., Using ETM+ and ASTER sensors toidentify iron occurrences in the Esfordi 1:100,000 mapping sheet of Central Iran.(2013), Journal of African Earth Sciences 85(0), 103-11410.1016/j.jafrearsci.2013.05.003Search in Google Scholar

[81] Lobo F.L., Costa M.P.F., and Novo E.M.L.M., Time-series analysis of Landsat-MSS/TM/OLI images over Amazonian waters impacted by gold miningactivities. (2015), Remote Sensing of Environment 157(0), 170-18410.1016/j.rse.2014.04.030Search in Google Scholar

Received: 2016-11-9
Accepted: 2017-6-12
Published Online: 2017-10-3

© 2017 Ke-Fa Zhou and Shan-Shan Wang

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Downloaded on 11.6.2024 from https://www.degruyter.com/document/doi/10.1515/geo-2017-0027/html
Scroll to top button