Study On Spectral Response To Vertical Movement Of The Metallogenic Elements Along Rock-Soil-Vegetation And Its Application

Hyperspectral remote sensing has been successfully applied to altered mineralextraction and other mineral exploration in sparsely vegetated area and bare area.Meanwhile the stress spectra detection of excessive heavy metal contents has beenused widely, and metallic elements mirgration within nanoscale have been detectedsuccessfully, which provide new ideas for mineral investigation and prediction byhyperspectral remote sensing in vegetated area.Supported by the project “rock week information extraction model and softwaredesign by satellite hyperspectral remote sensing”, and Huma, Heilongjiang is selectedas the study area. The vegetation coverage of forest and grassland in the area reachesas much as95%, and detritus layer and weathered layer are rather thick. Theemergence stratum is mainly formed in early Paleozoic Ordovician, Devonian andSilurian, and tectonic patters are composed of faults and folds mostly in the north-eastand north-west direction. Detailed geological survey and vegetation plot investigationwere carried out in August,2010. Nine sample points were selected, and the samplesfor the whole profile along rock, soil (Layer C, Layer B and Layer A) to vegetation(root, stem and leaf), and13elements contents are measured. To analyze themigration features of ore-forming elements in different growing periods of vegetationand their stress spectral response, the additional samples of vegetation are collected at9sample points in June, July and September of2011and their contents were tested.Only when the ore-forming elements are in active state in supergene conditionsthat the contents increase or decrease obviously, do the ore-forming elements result inabnormal contents. There exists obvious migration of elements, including Co, Zn, Bi,Mo, Au, Cu and so on after analyzing the elements from soil and rock at9spot points.Most element contents of soil-rock samples are higher than that of vegetation, for thevegetation grows in high supplies and high background. The soil, rock and vegetationsamples were crushed into power during the element testing and their spectra ofvegetation were collected in a dark room. The wavelet coefficients are calculated forthe whole profile of samples spectra by wavelet transformation. As the element contents of rock, soil and vegetation change, the wavelet coefficients show theopposite trend. There exists significantly negative relationship between waveletcoefficients and elements covering Co, Cu, Ni, Mo, Au, W, Pb and As. Amongdifferent wavelet basis, Daubechies series are strong correlation.Transfer coefficients are adopted for element migration within vegetation indifferent months. The elements are enriched differently for different organs ofvegetation. The elements are enriched in root in June and in leaves in August. Theelement contents of herb leaf are strongly correlated with soil layer A, but that ofwood plants have high correlation with soil layer B, C, which may relate to rootdevelopment. Red edge and absorption depth are calculated in different growingseasons and the results show that the red edges in June and September exhibits blueshift compared with July and August, which shows physiological cycle of plants thattend to thrive in June and grow normally in July, ripens in August and wither andyellow in September. The absorption depths differ in different months and areinfluenced by many factors.The multiple regression equation between absorption depths and elementcontents of plant leaf’s spectra are established, by vertical migration from soil tovegetation for ore-forming element and their spectra response. Through significancetest, the best prediction equation is determined and is applied to EO-1Hyperionimages, the enrichment information of ore-forming elements covering Co, Cu, Ni, Auand so on. The rock, soil and vegetation samples are collected in element enrichedarea and the element contents are measured, and the element contents of soil andvegetation in enrichment area and are higher than that in non-enrichment area.Geochemistry anomaly map of single element and element anomaly are extracted, Co,Cu, Ni have a good overlaying effect but Au is less relevant because of its minorcontent.