土地利用数据尺度转换的精度损失分析
|
摘要
由于栅格数据便于空间分析 ,因而通常将矢量数据转化成栅格数据来进行空间分析。在转化过程中 ,选用不同的栅格大小 ,其面积和精度损失是不同的。针对这个问题 ,以重庆市 1∶10万的土地利用矢量数据为例 ,探讨了不同栅格大小下 ,各种土地利用类型在转化过程的面积和精度损失。研究表明 :(1)栅格大小 <10 0m时 ,其精度损失均 <3 3%。 (2 )在 10 0 0m时 ,其精度损失达到 5 0 %以上的地类有 :湖泊、水库坑塘、滩地、农村居民点用地、工交建设用地、戈壁、裸岩石地、平原区旱地和坡度 >2 5°的旱地 ,这些地类的平均图斑大小均在 <5 1hm2 。 (3)在 30m～10 0 0m间 ,平均图斑大小 <82hm2 的地类随着栅格的由小变大 ,其面积变得比实际面积小 ;平均图斑大小 >10 1hm2的地类中除高盖度的草地和河渠之外 ,却与此相反
During spatial analysis, the vector data is always rasterized into the raster data because of its convenience for spatial analysis. How the area accuracy loss during rasterizing needs to be answered when the different grid size is adopted in order to choice a suitable grid size. The question is answered here by rasterizing the landuse vector data of Chongqing City with the grid of different size.Firstly,the landuse vector data is rasterisized with grid of different size such as 30m,60m,100m,200m,300m,400m,500m,600m,700m,800m and 1 000m.The value of each grid takes the landuse type code of the dominant area. Secondly,the area for each landuse type is calculated in grid.Thirdly,the area for each landuse type is also calculated in vector and used as the actual area.The number and average size of polygons for each landuse type are calculated.Finnaly,the loss of area and accuracy for each landuse type and each grid size is analyzed by comparing the area in grid with the area in vector.It is shown that (1) the area accuracy loss of all types of landuse is under 3.3% at the grid size under 100m. (2) the landuse types of the area accuracy loss above 50% include lake, pond, beach, rural residential area, traffic and building land, bare stone, plain dry land and dry land with slope above 25 degree at the grid size of 1 000m. Their average polygon size is less than 51 hm 2 . (3)The grid area of land use types with the average polygon size under 82 hm 2 except Gobi are less than their actual area with the increase of the grid size from 30m to 1 000m,But the grid area of land use types with average polygon size above 101 hm 2 except grass with high canopy and river are more than their actual area with the increase of the grid size from 30m to 1 000m.
During spatial analysis, the vector data is always rasterized into the raster data because of its convenience for spatial analysis. How the area accuracy loss during rasterizing needs to be answered when the different grid size is adopted in order to choice a suitable grid size. The question is answered here by rasterizing the landuse vector data of Chongqing City with the grid of different size.Firstly,the landuse vector data is rasterisized with grid of different size such as 30m,60m,100m,200m,300m,400m,500m,600m,700m,800m and 1 000m.The value of each grid takes the landuse type code of the dominant area. Secondly,the area for each landuse type is calculated in grid.Thirdly,the area for each landuse type is also calculated in vector and used as the actual area.The number and average size of polygons for each landuse type are calculated.Finnaly,the loss of area and accuracy for each landuse type and each grid size is analyzed by comparing the area in grid with the area in vector.It is shown that (1) the area accuracy loss of all types of landuse is under 3.3% at the grid size under 100m. (2) the landuse types of the area accuracy loss above 50% include lake, pond, beach, rural residential area, traffic and building land, bare stone, plain dry land and dry land with slope above 25 degree at the grid size of 1 000m. Their average polygon size is less than 51 hm 2 . (3)The grid area of land use types with the average polygon size under 82 hm 2 except Gobi are less than their actual area with the increase of the grid size from 30m to 1 000m,But the grid area of land use types with average polygon size above 101 hm 2 except grass with high canopy and river are more than their actual area with the increase of the grid size from 30m to 1 000m.
引文
[1]B .B .索恰瓦 ,等 .地理系统学说导论 [M] .北京 :商务印书馆 ,1981.
[2 ]NASA ,美国国家航空和宇航管理局地球系统科学委员会 (陈泮勤等译 ) .地球系统科学 [M] ,北京 :中国地震出版社 ,1992 .
[3]弗特普费尔 (江安宁译 ) .制图综合 [M ] .北京 :测绘出版社 ,1982 .
[4]CurtisE .WoodcockAlanH .Strahler .Thefactorofscaleinremotesensing[J] .Remotesensingofenvironment,1987,2 1:311～ 332 .
[5 ]HelmutMayerCarstenSteger.Scale_spaceeventsandtheirlinktoabstractionforroadextraction [J] .PhotogrammetricEngineering&RemoteSensing ,19985 3 :6 2～ 75 .
[6 ]HowaedVeregin .Quantifingpositionalerrorinducedbylinesimplification[J] .INT .J.GeographicalInformationSystem ,2 0 0 0 ,14(2 ) :113～ 130 .
[7]高志强 ,刘纪远 ,庄大方 .我国耕地面积重心及耕地生态背景质量的动态变化 [J] .自然资源学报 ,1998,13(1) :92～ 96 .
[8]高志强 ,纪远 ,庄大方 .中国土地资源生态环境质量状况分析[J] .自然资源学报 ,1999,14 (1) :93～ 97.
[9]黄裕婕 ,张增祥 ,周全斌 .西藏中部的生态环境综合评价 [J] .山地学报 ,2 0 0 0 ,18(4) :318～ 32 1.
[2 ]NASA ,美国国家航空和宇航管理局地球系统科学委员会 (陈泮勤等译 ) .地球系统科学 [M] ,北京 :中国地震出版社 ,1992 .
[3]弗特普费尔 (江安宁译 ) .制图综合 [M ] .北京 :测绘出版社 ,1982 .
[4]CurtisE .WoodcockAlanH .Strahler .Thefactorofscaleinremotesensing[J] .Remotesensingofenvironment,1987,2 1:311～ 332 .
[5 ]HelmutMayerCarstenSteger.Scale_spaceeventsandtheirlinktoabstractionforroadextraction [J] .PhotogrammetricEngineering&RemoteSensing ,19985 3 :6 2～ 75 .
[6 ]HowaedVeregin .Quantifingpositionalerrorinducedbylinesimplification[J] .INT .J.GeographicalInformationSystem ,2 0 0 0 ,14(2 ) :113～ 130 .
[7]高志强 ,刘纪远 ,庄大方 .我国耕地面积重心及耕地生态背景质量的动态变化 [J] .自然资源学报 ,1998,13(1) :92～ 96 .
[8]高志强 ,纪远 ,庄大方 .中国土地资源生态环境质量状况分析[J] .自然资源学报 ,1999,14 (1) :93～ 97.
[9]黄裕婕 ,张增祥 ,周全斌 .西藏中部的生态环境综合评价 [J] .山地学报 ,2 0 0 0 ,18(4) :318～ 32 1.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心