Dataset List

Vol.|Area

Oceanic region
Data Details

Google Earth Images Based Land Cover Data Validation Dataset for GlobeLand30 (2010) in the Region of Roof of the World


WANG Zhengxing1LIU Chuang2
1State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China2Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China

DOI:10.3974/geodb.2018.02.05.V1

Published:Mar. 2018

Visitors:321       Data Files Downloaded:66      
Data Downloaded:3.03 MB      Citations:

Key Words:

land Cover,GlobeLand30(2010),Roof of the World,validation Sample,accuracy assessment,Google Earth

Abstract:

GlobeLand30 (2010) is the first 30m spatial resolution global land cover dataset, which was produced by Chinese institutes led by National Geomatics Center of China. The global validation of GlobeLand30 (2010) was initiated by GEO in 2016 and served two purposes at least: an informative application of GlobeLand30 (2010) for users; and more efficient update of GlobeLand30 (2010) for producers. The validation in the region of Roof of the World (ROTW) is part of this global effort. Using Landscape Index Based (LSI) sampling method, 801 samples (plots) were extracted for ROTW, including: 24 plots for cultivated land, 127 plots for forestry, 331 plots for grass lands, 116 plots for shrub lands, 2 plots for wetland, 10 plots for water body, 1 plots for artificial land surfaces, 171 plots for bare land, and 19 plots for permanent snow and ice. Three steps were taken to label samples: (1) conversion of sample from point to 30 m x 30 m square; (2) conversion of shapefile to Google Earth format (.kmz); (3) labeling samples based on time serial Google Earth images. There is six major Fields in validation sample dataset files: the first three from Sampling period – SID, LCType, and LCName; the second three from Labeling period – CONF (confidence of labeling: 1 for good, sure for only one class, Class_1; 2 for fair, sure for one of two classes, Class_1 or Class_2; 3 for poorest), Class_1 (primary class), and Class_2 (secondary class, if CONF=2). The labeling results indicate that 725 out of 801 (91.51%) samples can be classified with confidence ‘good’ and ‘fair’ and can be used for accuracy assessment of GlobeLand30 (2010). The overall accuracy of GlobeLand30 (2010) ranges from 71.72% (if only Class_1 is used) to 83.86% (if both Class_1 and Class_2 are used). The dataset is archived in both .shp and .kmz data formats with the data size of 550 KB (Compressed to 141 KB in three data files).

Foundation Item:

Ministry of Science and Technology of P. R. China (2016YFA0600201, 2015DFA11360); Chinese Academy of Sciences (TSYJS04)

Data Citation:

WANG Zhengxing,LIU Chuang.Google Earth Images Based Land Cover Data Validation Dataset for GlobeLand30 (2010) in the Region of Roof of the World[DB/OL].Global Change Research Data Publishing & Repository,2018.DOI:10.3974/geodb.2018.02.05.V1.

Data Product:

ID Data Name Data Size Operation
1 Sample_LSI_ROTW_Point.kmz 30.42kb DownLoad
2 Sample_LSI_ROTW_Square.kmz 44.83kb DownLoad
3 ValPlotROTWGlobeLand30(2010).rar 66.28kb DownLoad
Co-Sponsors

Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences

The Geographical Society of China

Parteners

Committee on Data for Science and Technology (CODATA) Task Group on Preservation of and Access to Scientific and Technical Data in/for/with Developing Countries (PASTD)

Jomo Kenyatta University of Agriculture and Technology

Digital Linchao GeoMuseum