Study On Extraction Of Spectral Characteristics Of Vegetation And Inversion Of ChlorophyllContent In Reclaimed Land Of Ion-absorbed Rare Earth Ore

Using pool or heap leaching technology to exploit rare earth in ion-absorbed rare earth ore results in the change of soil characteristics of ore site,which results in the destruction of large areas of land and vegetation,and the environmental control mainly depends on artificial reclamation.However,because of the loss of soil nutrients caused by rare earth mining,and the lack of mineral components needed by the vegetation root system,the vegetation planted on the reclaimed land will not grow well,which will have a negative impact on the reclamation effect.According to the mechanism of visible spectrum of vegetation,it can be known that green peak and red edge are produced by chlorophyll,and chlorophyll can indicate the growth of vegetation and is an important index of vegetation growth estimation.There are more vegetation in ion-absorbed rare earth ore reclamation in South C hina,this paper takes typical representative bamboo willow,vernicia fordii tree and pinus elliottii as examples to extract the spectral information and chlorophyll and invert the chlorophyll content.It provides theoretical basis and technical support for vegetation growth monitoring in rare earth mining area.First of all,the original spectral data of vegetation were transformed by the first derivative transform and the envelope removal,analyzed and compared the spectral differences of the three vegetation.Secondly,chlorophyll sensitive bands were extracted by correlation analysis of chlorophyll content.Thirdly,combining trilateral parameter and vegetation index with chlorophyll content,the spectral characteristic parameters with good correlation are extracted.Finally,with the sensitive bands,trilateral parameters and vegetation index,using the univariate regression and stepwise regression method to build the retrieval model of chlorophyll content,and compare with the measured chlorophyll values.The results show that:?1?The original spectral green peak and red valley reflectivity of bamboo willow,vernicia fordii tree and pinus elliottii are higher than the general vegetation.The green peak moves 1⁴ nm to the shortwave direction,and the red valley moves 2¹0 nm towards the short wave direction.There is a phenomenon of"red shift"in different degrees.Compared with bamboo willow and pinus elliottii,the red edge position of the first derivative of vernicia fordii tree moves 23²4nm in the direction of short wave.With its growth situation,there is a phenomenon of"blue shift".?2?Based on the precision test of the inversion model of chlorophyll content of Bamboo willow,it was found that the stepwise regression model based on NBDI₆₄₂,D₆₇₁,BD₅₉₀90 and D₉₉₂92 show that the prediction R² is 0.823,RMSE is 2.915,and the RE is 0.0875.It can be used as the best inversion model of chlorophyll content of Bamboo Willow.It is found that the stepwise regression model based on?SD_r-SD_b?/?SD_r+SD_b?and?_v can be used as the best inversion model for the chlorophyll content of vernicia fordii tree.The R² is 0.830,RMSE is2.798,and the RE is 0.0778.For the pinus elliottii,the prediction results of stepwise regression model based on D₁₆₅₇,D₁₂₅₂ and D₄₈₀ are better than that of single variable inversion model.The R² is 0.847,RMSE is 2.153 and the RE is 0.0677.It can be the best inversion model for chlorophyll content.?3?The three kinds of plants are all stepwise regression models which have the best fitting and prediction effect,and the correlation between chlorophyll content and spectral characteristic parameters of different vegetation types was different.It shows that different vegetation adapts to environmental stress in ore district is different.The research results lay a foundation for monitoring the vegetation growth status in large area of ion-absorbed rare earth ore reclaimed land by hyperspectral technology,and also provide a certain basis for the formulation of reclamation measures.