Font Size: a A A

Study On Biomass Inversion And Scale Conversion Of Picea Schrenkiana Based On GF-1 And Landsat8 Image Data

Posted on:2020-12-29Degree:MasterType:Thesis
Country:ChinaCandidate:F XingFull Text:PDF
GTID:2493306602962719Subject:Ecology
Abstract/Summary:PDF Full Text Request
This research chooses the Picea schrenkiana of the central Tianshan Mountain as the study subject and utilizes the research date of the forests source sample plot,GF-1 and Landsat8 OLI images and DEM date of the research zone.Extracts the remote sensing factors of Picea schrenkiana biomass inversion with the help of RS and GIS methods,and then it analyses the relevance of Picea schrenkiana biomass with remote sensing factors.Builds the biomass remote sensing inversion model of Picea schrenkiana,then it estimates the Picea schrenkiana biomass with the model and analyses the spatial distribution characteristics of the Picea schrenkiana biomass.And based on the accurate comparison of Picea schrenkiana biomass inversion results in different scales,the Picea schrenkiana biomass conversion model between different spatial scale is established in this research.The main research results of the thesis are as follows:(1)The correlation analysis between Picea schrenkiana biomass and remote sensing factor demonstrates that the biggest correlation factors of Picea schrenkiana biomass are B4,B2,B1 and PCA1 which exist in single bands;the correlation coefficients are-0.360,-0.372,-0.346 and 0.388;DVI which exists in vegetation index,correlation coefficient is-0.341;CC3,ME3,VAR3,CO3,DI5,EN5,HO7 and SM7 which exist in image texture features,correlation coefficients are 0.560,-0.394,-0.307,-0.306,-0.371,-0.450,0.385 and 0.492.(2)The biomass remote sensing inversion model of the Picea schrenkiana as the follow:Biomass=2234.549+938.227*CO5-1114.888*DI5-132.861*ME3-1.016*B2-0.393B10.381*DVI+0.741*PCA1.The total biomass of Picea schrenkiana of the research area is 822365.33 t,average biomass is 85.78 t/hm2,with the accurate of 87.57%.The main area where biomass of Picea schrenkiana distributes is altitude from 1600 to 2800 m,slope form 5 degree to 45 degree,aspects of north,northwest and northeast.(3)The inversion accuracy of Picea schrenkiana biomass according to the GF-1 30 m image and Landsat8 OLI 30 m image are 85.61%and 76.38%,which decreased by 2.24%and 11.19%respectively compared with that from GF-1 2 m image.Taken the biomass inversion results of GF-1 30 m image and Landsat 8 OLI 30 m image as X and Y axes respectively,the scale conversion relationship between them was established as follows:Y=0.8999X+7.2138(R2=0.7781),the accuracy of biomass inversion results from Land8 OLI 30 m image was improved by 3.89%,reaching 80.27%,with the using of conversion relationship to calculate.The results of this study show that using GF-1 image data to inversion the Picea schrenkiana biomass has a high feasibility and with a accuracy higher than 85%.Compare to the central Tianshan Mountain,the height of distribution altitude of Picea schrenkiana biomass are higher than that in the west Tianshan Mountain,while the distribution slope and aspect of Picea schrenkiana biomass is the same as that in west Tianshan Mountain.By the scale conversion relationship of Picea schrenkiana biomass inversion results in different spatial scales,the accuracy of Picea schrenkiana biomass inversion result in lower spatial resolution scale can be improved.
Keywords/Search Tags:multi-resource remote sensing data, Picea schrenkiana, biomass inversion, multiple linear regression, scale conversion
PDF Full Text Request
Related items