A comparison of university of Maryland 1km land cover dataset and a land cover dataset in China

Land cover plays important roles in the understanding of the physical, chemical, biological and anthropological process in the earth system sciences. Land cover map at scales from local to global has been produced using the remote sensing data by visual interpretation or automatic classification methods during the past several decades. University of Maryland (UMd) land cover dataset is a global land cover dataset based on remote sensing method produced in recent years. The UMd approach employed a supervised decision tree method to classify global land cover types. This paper makes a comparison of this UMd land cover dataset with a Chinese land cover dataset. Firstly, we present a method to compare land cover datasets produced at different time based on invariant reliable land unit. Secondly we compare UMd land cover dataset with Chinese land cover dataset (CLCD) based on above method. Finally we analyze the possible factors affecting the differences among the land cover classification datasets. The comparison results demonstrate that most of the land surface in China was identified as different types in these two datasets. For example, UMd maps 51.1% of the deciduous needleleaf forest units in CLCD to the mixed forest. The classification scheme and method used in these datasets are the most likely reasons to explain the differences between them.

[1]  Jesslyn F. Brown,et al.  Development of a land-cover characteristics database for the conterminous U.S. , 1991 .

[2]  C. Justice,et al.  Characterization and classification of South American land cover types using satellite data , 1987 .

[3]  John R. G. Townshend,et al.  Continental land cover classification using meteorological satellite data , 1983 .

[4]  J. Townshend,et al.  African Land-Cover Classification Using Satellite Data , 1985, Science.

[5]  Jiyuan Liu,et al.  Study on spatial pattern of land-use change in China during 1995–2000 , 2003 .

[6]  Xiangming Xiao,et al.  Land-cover classification of China: Integrated analysis of AVHRR imagery and geophysical data , 2003 .

[7]  J. Townshend,et al.  Global land cover classifications at 8 km spatial resolution: The use of training data derived from Landsat imagery in decision tree classifiers , 1998 .

[8]  J. Townshend,et al.  Global land cover classi(cid:142) cation at 1 km spatial resolution using a classi(cid:142) cation tree approach , 2004 .

[9]  D. Roy,et al.  An overview of MODIS Land data processing and product status , 2002 .

[10]  D. Randall,et al.  A Revised Land Surface Parameterization (SiB2) for Atmospheric GCMS. Part I: Model Formulation , 1996 .

[11]  S. Running,et al.  8 – Generalization of a Forest Ecosystem Process Model for Other Biomes, BIOME-BGC, and an Application for Global-Scale Models , 1993 .

[12]  J. Townshend,et al.  NDVI-derived land cover classifications at a global scale , 1994 .

[13]  H. Mooney,et al.  Modeling the Exchanges of Energy, Water, and Carbon Between Continents and the Atmosphere , 1997, Science.