Font Size: a A A

A Research Of Spatial And Temporal Distribution Of Rhododendron In Macheng Based On Remote Sensing

Posted on:2017-05-12Degree:MasterType:Thesis
Country:ChinaCandidate:D Q LuoFull Text:PDF
GTID:2283330488968517Subject:Crop
Abstract/Summary:PDF Full Text Request
There are more and more tourists becoming interested in Rhododendron of Hubei Macheng Guifeng Mountain parkland because of its large distribution area and purely species. The parkland attracted more and more tourists in every flowering season. For the purpose of management and protection of Rhododendron resources and tourism resources, we made a preliminary research of spatial distribution characteristics and the areas’trend of Rhododendron in Guifeng Mountain which provided some evidence for ecology research and provided scientific basis for management and protection of Rhododendron resources.We made a preliminary analysis on spectral curve of Rhododendron, determined the separability of Rhododendron and other features and found that band 543 is the best combination. Then compared unsupervised and supervised classification methods. We collected distribution information of Rhododendron by maximum likelihood method in the years 2001,2002,2003,2007,2010. Analyzed the distribution area of Macheng Rhododendron in different altitude, slope, slope direction and different year by DEM data. Analyzed the factor which influenced the distribution of Rhododendron by the meteorological characteristics of each year.The results of the whole study are as follows:1, We can distinguish Rhododendron from other features in its flowering period(from April to early May). The spectral curve of Rhododendron embodied in band 1 to band 5 chiefly. The best band combination is R5, G4, B3.2, We used April 16th 2003 ETM data, in R5, G4, B3 band combination, summarized effective visual interpreting marks. We collected Rhododendron information by three conventional classification methods like IOSDATA unsupervised classification, maximum likelihood and support vector machine (SVM) supervised classification. By comparing the results, concluded that maximum likelihood is the most suitable choice for new hand, unsupervised classification is a very practical preliminary methods and SVM needs more prior knowledge. We used maximum likelihood to extract other years’Rhododendron information.3, The Rhododendron resources in Macheng are mainly distributed in the region with high altitude and steep slope, concentrated in Guifeng Mountain and Dabie Mountain’s others regions with altitude above 900 meters. Low altitude area had a small distribution. Rhododendron more easily distributed in the temperate zone on the side of mountain near the plains. Rhododendron growing rapidly in year 2001 to year 2003. By year 2003, there was almost 110,000 acres local Rhododendron. Influenced by meteorological factors Rhododendron reduced to less than 10,000 acres in year 2007. By year 2010, it restored to more than 90,000 acres.4, In three meteorological factors temperature, precipitation and sunshine, temperature is the most important factor, precipitation followed, sunshine has the minimal impact.By field verification, we ensured the accuracy of the classification information. By study of Rhododendron’s flowering period and phenology, we ensured the rationality of the results. The research of spatial and temporal distribution of Rhododendron in Guifeng Mountain provided strong scientific support for local Rhododendron resources’ exploitation, management and protection. At the same time, the research provided reference for the application in resources inventory of the remote sensing dynamic monitoring. It was important in the resources investigation and monitoring research in our country.
Keywords/Search Tags:Rhododendron, spectral curve, classification, dynamic monitoring, Guifeng Mountain
PDF Full Text Request
Related items