A Remote Sensing Biogeochemical Survey in and around the

Linghou Cu- polymetal Deposit, Southeastern China

Ling Han, Chengyan Du

*

School of Geology Engineering and Geomatics, Shaanxi Key Laboratory of Land Consolidation and Rehabilitation,

Chang’an University, Middle-section of Nan'er Huan Road, Xi’an, 710064, China.

Email: dcychdedu@163.com.

Keywords: The Linghou deposit, biogeochemistry, the OLI image, foliage

Abstract: Using the Landsat 8 imagery and the field- measured vegetation spectra, a biogeochemical survey in the

Linghou ore field of eastern China was conducted. Findings indicated that no vegetation anomalies were

observed in the 5 to 4 ratioing image. From the perspective of spectral feature, plants within the Linghou

diggings, especially at the tailings pond and wastewater pool, “grow worse” than at the peripheries,

although several individual exceptions behave against this trend. Thus we should not ignore the local

pollution which might be a valuable cue for the concealed ore- bodies or ecological destruction.

1 INTRODUCTION

Remote sensing technology relies mostly on the

capability of the sensor to register spectral

signatures and other geological features related to

mineral deposits (Cavalli et al., 2009). The

identification of sites with likely occurrence of

hydrothermal alteration is a positive clue for

existence of associated ore minerals. Nevertheless, it

is not providing a necessary spectral resolution to

identify specific minerals because of the broad band

configuration of TM/ETM+, particularly in SWIR.

The Linghou deposit surveyed in this paper is

located in northwestern Zhejiang province,

southeastern China, widely covered by vegetation

even in winter, but abundant in copper- polymetal

deposits. Because of the very dense cover of regolith

and vegetation, an alternative approach of survey

that reduces the effect of this thick coverage must be

developed in order to expose the alteration zones

and the anomalies associated with the concealed ore

body (Sabins, 1999).

According to (

Sims and Gamon, 2002), a series of

vegetal physiological and ecological variation, e.g.,

leaf temperature, chemical component, pigment and

water content, etc., may occur in response to ore

geochemical anomaly coming from rocks and soils

on which plants growth heavily relies. Normally

such a toxic action caused by metal poisoning of

plants and/or other ecological stresses is twofold: i)

The red- shift and/or blue- shift, e.g., chlorophyll

normally has a significant absorption within the red-

spectral region, while vegetal mesophyll has a

remarkable near infrared scattering, and the red-

edge appears in between. Generally those poisoned

leaves in gold ore- field have a bluely shifted red-

edge position (Kooistra et al., 2004). In addition,

spectral red- shift associated with a declined content

of leaf water due to contamination of heavy- metals

also occurs. ii) Relative to the healthy canopy,

poisoned vegetal leaves may always have higher

brightness in the images. Due to a very thick

sedimentary- vegetation cover in this area,

concealed rocks and deposits have become the focus

of mineral exploration; and in this regard, the study

on plant spectrum and its link to (environmental or

metallogenetic) geochemical anomalies may make a

particular contribution, although these

characteristics are just an indirect reflection of

mineral occurrence and environmental destruction.

From the above, the objective of this article is to

conduct a remote sensing data- driven

biogeochemical survey in Linghou ore field, eastern

China.

Han, L. and Du, C.

A Remote Sensing Biogeochemical Survey in and around the Linghou Cu- polymetal Deposit, Southeastern China.

In Proceedings of the International Workshop on Environment and Geoscience (IWEG 2018), pages 273-282

ISBN: 978-989-758-342-1

273

Figure 1: Geological sketch map of the Linghou polymetallic deposit.

2 GEOLOGICAL SETTING

The Linghou deposit (119°11′45″E, 29°29′30″N) is

located in the western Zhejiang province of China. It

is a medium- scale Cu- Pb- Zn deposit whose

genesis type is known as the hydrothermal reworked

deposit, as shown in Figure 1. The surrounding

strata of this deposit are mainly upper Devonian and

middle to upper Carboniferous along with the

general strike of NE- SW. There are two ore blocks

in Linghou: Songkengwu and Linghou (Tang et al.

2015). Typical calc- silicate (skarn) alteration can be

only found in the Songkengwu ore- field. Phyllic

alteration characterized by disseminated sericite and

quartz, with minor pyrite and chalcopyrite is well

developed within the granites. Silicification

characterized by quartz in ores or quartz + pyrite +

chalcopyrite veins in strata and granites is closely

linked to Cu- Au- Ag mineralization (Zhang et al.

2013). The carbonate alternation, characterized by

earlier recrystallization of dolomite + calcite

associated with the Cu- Pb- Zn minerali- zation, and

later calcite clusters in a hydrothermal calcite cave

near the ore- bodies, and calcite (± galena)

veins/veinlets across the ore and granites, is also

quite common.

3 DATA COLLECTION AND

PRETREATMENTS

The multispectral image was collected from Landsat

8 OLI_TRIS digital products (spatial resolution: 30

m) that are freely downloadable from

http://glovis.usgs.gov/. Its Row/Path is 40/119, the

time of acquisition is 14-04-2013. Later in Mar.

2016, a series of data were collected from vegetation

leaf samples in and around the mine area. The

sampling location is threefold: 1, sampling within

the mine excavations; 2, sampling within the

peripheral or background area relative to diggings,

acting as a control group; 3, sampling in one or more

kindred deposit (s) nearby in order to make a

contrast. Using FieldSpec®4 HI- RES spectrograph,

the reflectivity curve of dozens of leaf samples were

tested in- situ. The spectral measurement of foliage

adhered strictly to the user’s guide of this apparatus.

The basic approach of data quality control is to do

radiometric calibration by keeping the probe

vertically orienting the horizontally placed

whiteboard which is an approximate Lambertian

reflector with a fixed and known reflectivity

(Hatchell, 1999). Particularly, if the weather

IWEG 2018 - International Workshop on Environment and Geoscience

274

condition was not as good as expected, the

frequency of calibration must be doubled— every

two minutes. Conversely, it would be the best if a

100% reference line of reflectivity could be obtained

in sunny days.

Several pre- treatments, like imagery cutting,

radiometric calibration, atmospheric correction, and

orthorectification were carried out. At the same time,

waters with NDVI (Normalized Difference

Vegetation Index) between -0.079 and 0.05 were

erased for subsequent analysis, which accounts for

about 27.19% of the total area. There is no need to

mask other objects like settlements,

(concrete/asphalt) roads, factory buildings, etc., for

they are too small to extract and have limited

influence on the analysis.

4 SPECTRAL ANALYSIS OF

VEGETATION COVER

4.1 Spectral Analysis

(Zhao et al. 2017) discovered that on the 1: 200000

stream- sediments geochemical map, there is a

significant copper (Cu) - polymetal anomalous zone

lying in and around this studied area, e.g., copper

content in its concentration center (119°11′47.88″E,

29°29′05.07″N) reaches up to 3358×10-6 ppm,

nearly 120 times higher than copper clarke number

(Rudnick and Gao, 2003). According to (Gan and

Wang 2004), a significant decline of chlorophyll

pigment is the most obvious visual sign of

vegetation poisoned by ore metals. In this respect, it

is believed that the band combination 5, 4 and 3 of

Landsat 8 OLI is typically applied for vegetation

analysis, and seemingly, intense red color represents

vigorous growing plants producing a lot of

chlorophyll, while lighter shades of reds are still

vegetation, but may either be mature trees/plants or

dead, unhealthy plants. Unfortunately, the composite

results seem not so ideal: most light red shades are

actually montanic shadowed outcrops, or to a lesser

extent sparse vegetation areas (Ghasemloo et al.,

2011). A similar scenario can apply to the NDVI or

band 5 to 4 ratioing imagery.

In the visible wavelengths, there are several

peaks of spectral absorption and reflection (e.g. a

blue trough at ~480 nm, a red trough at ~680 nm,

and a green peak at ~550 nm), sensitively associated

with foliage content of chlorophyll. The OLI band 5

at NIR has a striking platform of reflection, relative

to any other peaks of reflection between 1400 and

2500 nm, whose average reflection is normally

greater than or equal to ~70% and related to

moisture content. Inasmuch the 5 to 4 ratioing OLI

image should play a role extracting biogeochemical

anomalous information relating to metal toxic action.

In SWIR: OLI band 6 (1560~1660 nm) and band 7

(2100~2300 nm) are quite sensitive to foliage

moisture content and mesophyll cellular structure.

Figure 2 shows the darkest pixels in the OLI 5/4

ratio image, which, in theory, represents the

unhealthy vegetation. But in reality, they are just a

reflection of hill- shaded areas.

According to (Tong 1988), the Linghou mine is

considered to be a sedimentary- transformed

stratabound type. The syngenetic sediments may be

rich in Cu and other ore metals, but far from mineral

industrial grade. The Cu- polymetal- enriched sandy

shale layer is located mainly in -50 m and deeper

underground.(Cheng et al. 2012) pointed out that the

covering layer plays an essential role in shielding

and attenuating ore elemental migration, so that the

deep- rooted geochemical anomalies and metal-

logenesis exhibit only gentle and low grade

anomalies at the surface. However, due to a special

geochemical property, it is reported that Cu

abundance in soils at the upper part of the copper

deposit comes up to n000 ppm, which can be used

for secondary- halo prospecting (Liu et al., 1980).

Also local edaphic geochemical or vegetal

anomalies related to ore (Jiang et al., 2002) in

Linghou mine were confirmed (Yang et al., 2009).

Table 1, e.g., shows that the average abundance of

Cu in the Linghou diggings is significantly greater

than normal 5 ppm. As the relevant anomalies

cannot be exposed in Figure 2, so we have to turn

our focus to local and limited stressing phenomena

related to mine pollution, which is probably beyond

the monitoring ability of OLI imagery.

4.2 Vegetational Hyperspectral

Analysis

Although the vegetational coverage severely

masks/alters surface signals related to metallogenic

alteration and environmental deterioration, we

cannot exclude the possibility that by feat of spectral

sampling under similar illumination conditions,

vegetation information affected more or less by

various stressing agencies, and hence could act as an

indicator serving to get some specific anomalies

A Remote Sensing Biogeochemical Survey in and around the Linghou Cu- polymetal Deposit, Southeastern China

275

associated with mineralization in this area (Ma, 2000).

Table 1: The biogeochemical anomalies (concentrations of Cu and Zn in the plant leaves) in the Linghou and adjacent

Fuyang mining areas

a

.

Site Species Cu (mg kg

–1

) Zn (mg kg

–1

)

Residue sand place

Polygonum lapathifolium 150.2 505.6

Alternanthera sessilis 220.3 280.2

Eclipta prostrata 125.7 525.6

Farmland

Siegesbeckia glabrescens 190.1 95.4

Eclipta prostrata 235.4 750.4

Old mining area

Elsholtzia splendens 687.0 530.3

Rumex acetosa 530.0 530.3

Artemisia lavandulaefolia 198.6 372.5

Viola diffusa 606.4 585.6

a

After (Jiang et al., 2002).

Figure 2: The darkest colored areas in the OLI 5/4 ratio image,

namely the so- called vegetation anomalies. The relevant

threshold was determined by visual interpretation.

Figure 3: The sampling position P-2 at the periphery

of mine (29°29′16.56′′N, 119°11′ 41.76′′E). The main

lithology in this profile is carbonate- rock formation

(C2h), within this limestone wall no conspicuous

quartz mines and granitic apophyses were found. Here

the withered Miscanthus floridulus leaves were

collected for spectral analysis.

4.2.1 Phyllostachys Pubescens

Phyllostachys pubescen is one of the most important

cash crop in western Zhejiang province, but most

phyllostachys pubescens in later March are not in

their vigorous growth period, and many leaf edges

look still withered yellow. Four samples of

phyllostachys pubescen leaves were collected in

three sites. Two of them are at the mine (noted as PP

at O-1, and O-2), while the other two acted as the

comparative samples, as shown in Figure 3, are at

the periphery of mine (PP at P-1, and P-2). P-1 is

closer to the diggings than P-2. O-1 is at a small

catchment basin (29°29′17.04″N, 119°12′5.69″E),

and it was sampled close to the surface runoffs,

ensuring that the plants root system nearby can

absorb enough ore metals related to mineralization.

O-2 was collected at 29°29′21.19″N,

119°11′52.34″E. P-1 was tested at 29°29′21.06″N,

119°12′4.87″E. This site is on a hillside adjacent to

the mining area, and there is no surface runoff, ore-

bearing quartz veins and outcropped bed- rock

IWEG 2018 - International Workshop on Environment and Geoscience

276

nearby. These samples are fresh and plausibly

healthy leafage, leaving out partly withered or newly

transplanted plants. The weather condition during

test was cloudy to sunny.

As seen in Figure 4, the reflectance curves of

these 4 samples are similar in shape but distinct in

numerical magnitude, so spectral ratioing (or

derivative) is the best way to eliminate these effects.

For example, the average of foliage reflectance

between 850 and 880 nm, which is just the spectral

region of band 5 of an OLI image, is noted by B5

here; while the average between 640 nm and 670 nm,

namely the spectral region of band 4, is noted as B4

here; and the ratios of B5 to B4 of PP at O-1 is 8.12,

O-2 is 7.18, P-1 is 10.78, and P-2 is 6.84.

Apparently, the growth conditions of O-1 and O-2 in

diggings site are inferior to P-1; the phyllostachys

pubescen tree at P-2, however, grows in a little pile

of regosol within a small limestone sink- hole; so it

seems that the subalimentation, not the biotic stress

associated with ore contamination, may be

responsible for its lower value of B5/B4, which is in

accordance with (Jiang et al. 2013) report that

calcareous soils are characterized by

neutral/slightly- alkaline chemistry, thinner soil

layer, lack of available moisture, and etc., and few

wood species in them could grow well. Additionally,

(Li 2007) pointed out that due to polymetal stressing,

foliage reflection within 550~680 nm must be

enhanced significantly, that is why the reflectivity of

PP at P-1 and P-2 seems low.

Figure 5 further demonstrates the 1st order

derivative reflectivity between 680 and 750 nm: the

derivative of PP at P-1 reflectivity peaks at 719 nm,

known as the red- edge (Rock et al., 1988). Relative

to P-1, there seems a significant blue- shift of PP

red- edges at O-1 (701 nm) and O-2 (702 nm). PP

red- edge at P-2 is 700 nm. The slopes of the red-

edge are 0.0073 (O-1), 0.01093 (O-2), 0.005344 (P-

1) and 0.008896 (P-2), respectively, and PP at P-1

therein is the most gentle one. Actually, variation in

chlorophyll and carotenoid concentrations, in

addition to a buildup of additional pigments such as

tannins, may occur within leaves in response to

stress like heavy- metal pollution in the mining area,

and a blue- shift phenomenon may always act as a

diagnostic signature helping to identify the

biogeochemical damage caused by metallic stress

agents (Ninomiya et al., 2005). However, due to a

relatively small amount of spectral shifts (versus a

relatively larger spectral resolution of the

spectrograph), the observation error of discrete

spectra, and the spectrum rebuilding error, the red-

edge position are not so accurate sometimes, and

thus the utilization of this parameter must be

considered with cautions.

Another important parameter describing

vegetational anomalies is the so called “band-

depth”. As displayed in Figure 6- a, taken a foliage

sample at O-1 as instance, the reflectance with

continuum removed is noted as R’, and the band-

depth (D) at each point is calculated by D = 1 - R’

(Sanches et al., 2014). Meanwhile, in order to

eliminate the effects of topography, landforms, and

atmosphere, and for ease of comparison, normalized

band- depth Dn was further worked out by Dn = D /

Dc, where Dc is the band- depth of the central band,

which, in this work, is assigned the maximum of D

within a wavelength range of 350 to 1500 nm. In

this way, a continuum- removed reflectance curve is

normalized to a range of 0~1.0 that can enhance the

vegetal absorption and reflection spectrum. In

Figure 6, only the first three peaks of absorption are

taken into account in this paper, the first two

appearing in visible region: the blue trough and the

red trough, and the other one developing in near-

infrared range and reflecting foliage moisture.

Because of the blue- or red- shift possibly existing,

the depths and position of these peaks may not be

fixed.

As listed in Table 2, for the first peak, the band-

depths of PP at O-1, O-2, and P-2 are significantly

greater than at P-1, implying a deepened trough of

absorption; also, there is a blue- shift of PP from P-1

to O-2 and P-2. For the third peak, PP at O-2 and P-

2, relative to P-1, seem shifted bluewards, and actual

band- depths of O-1, O-2, and P-2 are much deeper

than P-1. This feature seems not completely

consistent with the phenomena that foliage badly

stressed by metals may have a saliently shallowed

band- depth.

A Remote Sensing Biogeochemical Survey in and around the Linghou Cu- polymetal Deposit, Southeastern China

277

Figure 4: Spectral curves of reflectance of four phyllostachys

pubescen leafage Samples. Note that there is a drastic

fluctuation around 1900 nm, which may have something to do

with the illumination conditions and the surroundings like

breeze.

Figure 5: The 1st- order derivative reflectivity

(680~750nm) of 4 phyllostachys pubescen samples.

Figure 6: (a) The reflectivity curve of sample O-1, the red

curve: initial reflectivity curve (for both (a) and (b)), and

the green one: the continuum- removed curve of

reflectivity.

Figure 6: (b) Normalized band- depth of O-1 (the black

line).

Figure 7: Spectral curves of reflectance of four fir leafage

samples, and the corresponding sample description is seen

in the context.

Figure 8: The central position of sampling sites of O-1, O-

2 and O-3.

The site of O-3

The site of O-1, nearby

a mine- wastewater pool

The site of O-2,

nearby a dressing

IWEG 2018 - International Workshop on Environment and Geoscience

278

a

Note: The average of foliage hyperspectral reflectance between 850 and 880 nm is noted as band 5 (also the spectral range of Band 5 of an

OLI image), while the average between 640 and 670 nm, namely the spectral range of Band 4 of an OLI image, is noted as B4, and the

“band- ratioing 5/4” is actually a ratio of B5 to B4, no spectral anomalies can be observed in the image; or the relevant anomlies are

submerged into the hill- shaded areas, and hence indistinguishable. O-1, 2, 3 and P-1, 2 stand for different sampling sites, see context for

the positions for O-1, O-2, P-1 and P-2. O-3 is at 29°2 9′21.06′′N, 119°12′4.87′′E. WL is short for “Wave- length”, and NU is short for

“name known”, NU-1 therein is a newly grown leaf. Due to my limited professional knowledge on phytology, some samples were unable

to be identified in the open air, most of them, however, were collected from local arbor species. MF- miscanthus floridulus, IC- incense

cedar, OT- osmanthus tree, MP- masson pine, PP- phyllostachys pubescen, CT- camphor tree, M- moss, F- Chinese fir. Besides, peak1:

between 350nm and 557nm, the maximum value of normalized band- depth and the wavelength it corresponding to; peak 2: between 558

and 750, the maximum of band- depth normalized and the wavelength; peak 3: between 1400 and 1500nm, the maximum band- depth and

its position of wavelength.

Table 2: The band- depth at three troughs of absorption of all the foliage samples in Linghou mine

a

.

Site Species

Band- Depth

“Band 5/4”

The red- edge

WL peak-1 WL peak-2 WL peak-3 Slope WL

G

R

O

U

P

I

0-3 NU-1 550 0.948 759 1 1650 1.051 1.164 0.002 720

0-3 MF 490 0.975 678 1 1454 1.011 3.481 0.013 696

0-3 IC 502 0.986 676 1 1453 1.006 3.706 0.004 719

0-2 M 498 1 678 1 1459 1.006 5.048 0.007 695, 696

0-1 MF 498 0.957 679 1 1368 1.035 5.292 0.008 721

P-1 Arbutus 505 0.999 680 1 1457 1 5.529 0.006 719

0-2 M 440 0.992 679 1 1459 1.002 5.665 0.009 696

0-1 MF 494 0.987 675 1 1456 0.999 6.423 0.023 719

P-2 PP 493 0.981 675 1 1454 0.990 6.742 0.009 700

0-3 OT 497 0.983 675 1 1458 1.001 6.925 0.015 719

0-3 PP 501 0.977 678 1 1458 0.992 7.102 0.011 701, 702

G

R

O

U

P

Ⅱ

Site Species WL peak-1 WL peak-2 WL peak-3 “Band 5/4” Slope WL

0-1 PP 499 0.933 677 1 1461 0.987 7.727 0.007 701, 702

0-2 NU-2 496 0.977 678 1 1460 1.002 8.062 0.020 701

0-1 OT 502 0.994 675 1 1463 1.001 8.870 0.010 719

0-2 F 495 0.978 675 1 1459 1.001 8.956 0.010 719

0-1 F 497 0.999 672 1 1356 1.063 9.089 0.011 719

0-2 NU-3 496 0.992 676 1 1452 1 9.497 0.013 719

0-3 MP 496 0.984 673 1 1458 1.004 9.810 0.014 719

P-1 PP 500 0.890 673 1 1460 0.892 10.650 0.005 719, 725

P-1 F 492 0.998 675 1 1453 1 10.851 0.007 719

0-1 F 498 0.988 674 1 1360 1.511 10.852 0.010 721

0-3 NU-4 498 0.994 674 1 1456 0.999 10.922 0.014 701

0-1 NU-5 498 0.994 673 1 1455 1 11.070 0.013 719

0-2 CT 499 0.973 676 1 1460 0.990 11.115 0.013 719

P-1 OT 502 0.991 678 1 1457 1.003 11.473 0.012 719

0-1 MF 498 0.999 671 1 1399 1.408 11.698 0.005 719, 720

0-3 NU-6 502 0.986 672 1 1454 1 12.092 0.013 719

0-3 NU-7 490 0.992 675 1 1455 1.001 12.846 0.011 719

G

R

O

U

P

Ⅲ

Site Species WL peak-1 WL peak-2 WL peak-3 “Band 5/4” Slope WL

0-1 Loquat 494 0.952 673 1 1360 4.794 18.629 0.009 719

P-2 Loquat 497 0.987 674 1 1454 0.990 18.781 0.011 719

P-1 Arbutus 498 0.997 676 1 1449 1 19.190 0.002 721

P-2 OT 497 0.980 673 1 1451 0.989 19.481 0.014 719

P-1 Citrus 500 0.995 669 1 1457 0.997 20.393 0.014 719

A Remote Sensing Biogeochemical Survey in and around the Linghou Cu- polymetal Deposit, Southeastern China

279

4.2.2 Cunninghamia Lanceolate

Cunninghamia lanceolate is commonly called

Chinese fir. Evergreen plants can retain chlorophyll

even under freezing conditions (Yokono et al., 2008),

so fir samples here may provide a better spectral

comparison. Four foliage samples of Chinese fir

were collected. The first two were sampled at O-1,

noted by F at O-1-1 and O-1-2, the third one is noted

by F at O-2, and the fourth one, noted by P-1, was

tested in the periphery of mine. As shown in Figure

7, P-1 of Chinese fir has the lowest reflectivity as

expected; and “band ratios 5/4” of F at O-1-1

(10.9757), O-1-2 (9.21867), and O-2 (9.086264) are

simultaneously less than P-1 (11.0689). Between

650 and 750 nm, a shift of the steep slope to shorter

wave- lengths (blue- shift) is seen in reflectance data

of O-1-1, O-1-2, and O-2 relative to P-1; while

between 1250 and 1450 nm, it shifts towards the

longer wave length direction (red- shift). There is no

obvious change of the red- edge inflection point of

these fir samples, while the slope of the red- edge of

F at P-1 is gentler (0.007).(Ma 2000) had related

changes in the slope of the red- edge (700~740 nm)

to chlorophyll concentrations in the foliage, and

both the position and slope of the red- edge will

change as healthy leaves progress from active

photosynthesis through metals- induced stresses and

various stages of senescence (natural decline) due to

loss of chlorophyll and the addition of tannins.

In addition to the curves of normalized band-

depth from 350 nm to 1500 nm, for the first peak,

there is a significant red- shift of F at O-1-1 (498

nm), O-1-2 (497 nm) and O-2 (495 nm), relative to

P-1 (491 nm); meanwhile, F at P-1 has a slightly

greater and lower value of band- depth (0.996181,

Dc here used for normalization is assigned the

maximum of band- depths within 350 to 1200 nm)

than the other three: O-1-1: 0.977587, O-1-2:

0.998518, O-2: 0.955699; for the second peak, there

are blue- shifts of O-1-1 (674/675 nm), O-1-2 (673

nm) and O-2 (676 nm), relative to P-1 (677 nm); for

the third peak, F at O-1-1 and O-1-2 are ignored,

while O-2 (1459 nm) relative to P-1 (1452 nm)

appears a 5 nm red- shift, but the actual band- depths

of them have little difference. Like phyllostachys

pubescen (PP), these features are basically

consistent with the phenomena of foliage poisoned

more or less by ore heavy- metals (Xu et al., 2003).

4.2.3 Other Plants

From Table 2 which is a list of spectral parameters

of some other foliage samples collected at Linghou,

We noticed that there are many individual

exceptions behaving against the general pattern

summarized above, whereas the “ratioing 5/4”

(B5/B4) seems relatively stable, and have a

characteristic feature dividing original foliage

samples into several distinctive groups, e.g., lower

damage site and higher damage site (Rock et al.,

1988), although, admittedly different species may

correspond to different “ratioing 5/4” values as seen

in Table 2.

As manifested in Table 2, the first group

includes the first 11 samples within the diggings,

adding two anomalous samples at the periphery. The

second group is composed of fifteen samples, three

of them, i.e., phyllostachys pubescen, fir, and

Osmanthus tree, were collected outside the mine,

and the remainders are all from the mining district.

The relatively healthy third- one has five samples;

four of them were sampled at P-1 and P-2, adding

one sample of loquat from O-1. As exhibited in

Figure 8, the sites O-1, 2, and 3 were sampled

nearby the contaminated zones, so the spectral

anomalies, as a response to trace metals stressing, in

group I and II must be an inevitable phenomenon.

Fortunately, parametric characteristics of several

leaf samples in group II and III imply that severe

destruction of ecosystem in the Linghou ore deposit

which is adjacent to the famous Qiandaohu

International Tourism Area seems limited in scale

and slighter in degree, identified with the findings

revealed in the OLI multispectral image. On average,

leaf samples within the diggings, as a whole, do

have a relatively lower B5/B4 ratio: the average of

“ratio 5/4” of peripheral samples (P-1 and P-2) 13.7

is much greater than that of the samples within the

mining district (O-1, -2 and -3) 8.6. Likewise, the

average of spectral reflectivity between 350 and

1300 nm may also act as a diagnostic parameter

reflecting the overall degree of metals stressing, and

we found that the average value of this parameter of

peripheral samples 0.34 is still a bit greater than that

of the samples at the mine 0.26.

Table 2 further reads that: for peak 1, the average

band- depth in group I is 0.980455, in group II is

0.980118, and group III 0.9822. Obviously, the

depths of absorption in group I and II becomes

shallower. For peak 2, the average wavelength

position of Group I, Group II and Group III is

IWEG 2018 - International Workshop on Environment and Geoscience

280

684.73 nm, 674.53 nm and 673 nm, respectively.

From I to II, and III, there is an obvious blue- shift.

The average wavelength position of the red- edge of

foliage in group I is at 709.6 nm, in group II is 716

nm, and in group III 719.4nm, this is a blue- shift

process also, in accord with the spectral

characteristics of vegetation under metal stressing.

For peak 3, the median of the wavelength position of

samples in group I is 1457 nm, in group II is 1456

nm, and group III 1451 nm, indicating a red- shift.

Herein the substitution of average for median is

operated for suppressing the drastic fluctuations of

the reflectivity curve (appearing from time to time)

from 1400 to 1500 nm. Moreover, taking advantage

of the OLI image to perform vegetational pollution

monitoring, the parameter band- ratioing 5/4 (or

NDVI) has merely a 30 m spatial resolution, namely

the spectral feature of mixed pixels within a 30 m ×

30 m grid, and that is why the statistic mean method

was used here to analyze Table 2. Nevertheless,

there are many individual exceptions behaving

against the overall pattern summarized above.

5 DISCUSSIONS

Although the stress agents responsible for the

observed spectral characteristics have not been

identified by testing of foliage biochemical

components, mine wastewater and surficial tailings

ponds do exist within the diggings.(Song et al. 2014)

believed that anomaly of RS biogeochemistry has

obvious internal identity with soil chemical anomaly

in material quantity and quality, although vegetal

self- protective mechanism decreasing the

absorption of ore metals may be activated. Actually,

even in and around the world- class Dexing Mine

that is 160km away from Linghou, not every foliage

species is metals- enriched, even different parts of

one individuality have distinctive metal content

(Jiang et al., 2013), let alone those individual

exceptions of different tree species sensitive to

different metals as well as different sampling sites

suffering different degrees of pollution. In Table 2,

e.g. Osmanthus foliage at P1 was sampled at a small

hillside, while its counterpart at P2 was collected in

the agrarian zone, the latter, unlike PP at P-2, has a

bit larger 5/4 “ratioing”, and the loquat sample’s

ratio at P1 and P2 seems similar. Therefore sample-

grouping and group- averaging aforementioned are

essential (Zhao et al., 2017). Thus, we practically

reconfirmed poisoning metals stressing on local

vegetation at Linghou perhaps is too weak to form a

characteristic fractal dimension or a specific

normally distributed mass by itself, standing out

from the drastic difference of brightness and shade

hue caused by ruggedness of relief everywhere, but

we still need to keep an eye on those small- scaled

prospecting and pollutional criteria revealed by plant

spectral abnormality.

6 CONCLUSIONS

It is certainly true from the remote sensing data that

spectral anomalies can be observed in several leaf

samples, especially those collected close to the

tailings ponds and the wastewater pools.

With a few exceptions, most of the known

samples do not show diagnostic spectral anomalies.

For those “exceptional” samples, it is impossible to

determine whether the relevant anomalies are related

to the ore body and the associated soil heavy- metal

contamination, or to other stressing agencies

because the leaf biogeochemical measured data is

missing.

The biogeochemical remote sensing technology

is rarely available in the ore prediction and

environment pollution estimation, if there is no

display of extensive spectral anomalies in the image.

This article provided a counterexample that the

biogeochemical RS information should be treated

with great caution. The so- called RS

biogeochemical anomalies might be untenable, if no

further evidence, such as the measured leaf heavy-

metal content, can verify their existence. Such

anomalies are better to be considered as an auxiliary

approach for ore prediction or environmental

monitoring, and it provides a good idea of where

further research should be on a specific area instead

of wasting resources in places where the possibility

of any positive findings is very slim. This is what

our research is about.

ACKNOWLEDGEMENTS

This work was financially supported by the 1:50,

000 geological mapping in the loess covered region

of the map sheets: Caobizhen (I48E008021),

Liangting (I48E008022), Zhaoxian (I48E008023),

Qianyang (I48E009021), Fengxiang (I48E009022),

& Yaojiagou (I48E009023) in Shaanxi Province,

A Remote Sensing Biogeochemical Survey in and around the Linghou Cu- polymetal Deposit, Southeastern China

281

China, under Grant [DD-20160060]. And the project

of open fund for key laboratory of land and

resources degenerate and unused land remediation,

under Grant [SXDJ2017-7].

REFERENCES

Cavalli R M., Laneve G and Fusilli L, et al. 2009 Remote

sensing water observation for supporting lake victoria

weed management Journal of Environmental

Management 90 7

Cheng Q 2012 Ideas and methods for comprehensive

prediction of mineral deposits in covered areas Earth

Sciences 37 06

Gan F P and Wang R S 2004 Research of the

technological methods of extraction of remote- sensing

rock and mineral information Beijing: Geological

Publishing House (In Chinese)

Ghasemloo N, Mobasheri M R and Rezaei Y 2011

Vegetation species determination using spectral

characteristics and artificial neural network (scann)

Journal of Agricultural Science and Technology 13 4

Hatchell D C 1999 ASD Technical Guide. 3rd Edition,

Analytical Spectral Devices Inc, Boulder

Jiang L Y, Shi W Y and Yang X E, et al. 2002 Cu-

hyperaccumulators in mining area Chinese Journal of

Applied Ecology 13 7 (In Chinese with English

abstract)

Jiang M J, Hu J Q and Si W J 2013 A survey and filtration

of copper- enriched plants in Dexing copper mining

zone Jiangxi province Jiangsu Agricultural Sciences

41 7(In Chinese)

Kooistra L, Salas E A L and Clevers J G P W, et al. 2004

Exploring field vegetation reflectance as an indicator

of soil contamination in river floodplains

Environmental Pollution 127 2

Li N 2007 The spectral and image characteristics of

vegetation in the press of heavy metal Master’s

dissertation of Shandong University of Science and

Technology (In Chinese with English abstract)

Liu Y J, Cao L M, Li Z L, et al. 1980 Elements

geochemistry. Beijing: Science Press (In Chinese)

Ma Y L 2000 Biogeochemical characteristics of the Hetai

gold deposit, Guangdong province and their

significance in remote- sensing exploration Acta

Mineralogica Sinica 20 1 (In Chinese with English

abstract)

Ninomiya Y, Fu B and Cudahy T J 2005 Detecting

lithology with advanced spaceborne thermal emission

and reflection radiometer (ASTER) multispectral

thermal infrared “Radiance- at- Sensor” data Remote

Sensing of Environment 99 1-2

Rock B N, Hoshizaki T and Miller J R 1988 Comparison

of in situ, and airborne spectral measurements of the

blue shift associated with forest decline Remote

Sensing of Environment 24 1

Rudnick R L and Gao S 2003 Composition of the

continental crust Treatise on geochemistry 3

Sabins F F 1999 Remote sensing for mineral exploration

oregeol rev Ore Geology Reviews 14 3-4

Sanches I D, Filho C R S and Kokaly R F 2014

Spectroscopic remote sensing of plant stress at leaf and

canopy levels using the chlorophyll 680nm absorption

feature with continuum removal Isprs Journal of

Photogrammetry and Remote Sensing 97

Sims D A and Gamon J A 2002 Relationships between

leaf pigment content and spectral reflectance across a

wide range of species, leaf structures and

developmental stages Remote Sensing of Environment

81 2

Song C A, Song W and Lei L Q, et al. 2014 Effect of plant

physiological ecology factors on the formation of

metalliferous deposit biogeochemical anomaly

Advances in Geosciences 04 06 (In Chinese with

English abstract)

Tang Y, Li X and Zhang X, et al. 2015 Some new data on

the genesis of the linghou Cu- Pb- Zn polymetallic

deposit—based on the study of fluid inclusions and C-

H- O- S- Pb isotopes Ore Geology Reviews 71 (In

Chinese with English abstract)

Tong M X 1988 Minerogenetic mechanism and potential

ore- bearing analysis—No.2 ore body of Jiande Copper

deposit as an example Geology of Zhejiang 4 2 (In

Chinese with English abstract)

Xu R S, et al. 2003 Remote sensing biogeochemistry

Guangzhou: Guangdong publishing house of science

and technology (In Chinese)

Yang M G, Huang S B and Lou F S 2009 Lithospheric

structure and large- scale metallogenic process in

Southeast China continental area. Geol. China 36 (In

Chinese with English abstract)

Yokono M, Akimoto S and Tanaka A 2008 Seasonal

changes of excitation energy transfer and thylakoid

stacking in the evergreen tree taxuscuspidata : how

does it divert excess energy from photosynthetic

reaction center? Biochimica Et Biophysica Acta1777 4

Zhang D H, Wang K Q and Zhu Y D 2013 Project paper

of “regionally metallogenic regularity and prediction of

copper- polymetal deposits in the west- middle section

of the Jiang- Shao collision- orogenic belt” Beijing:

China University of Geosciences (Beijing) (In Chinese)

Zhao B and Han L, et al. 2017 Ore- and bio- geochemical

survey based on the landsat remotely sensed data in

and around the Dexing porphyry copper- polymetal

ore- field, southeastern China Journal of the Indian

Society of Remote Sensing 7

IWEG 2018 - International Workshop on Environment and Geoscience

282