Utilizing the global land cover 2000 reference dataset for a comparative accuracy assessment of global 1 km land cover maps

Many investigators use global land cover (GLC) maps for different purposes, such as an input for global climate models. The current
GLC maps used for such purposes are based on different remote sensing data, methodologies and legends. Consequently,
comparison of GLC maps is difficult and information about their relative utility is limited. The objective of this study is to analyse
and compare the thematic accuracies of GLC maps (i.e., IGBP-DISCover, UMD, MODIS, GLC2000 and SYNMAP) at 1 km
resolutions by (a) re-analysing the GLC2000 reference dataset, (b) applying a generalized GLC legend and (c) comparing their
thematic accuracies at different homogeneity levels. The accuracy assessment was based on the GLC2000 reference dataset with
1253 samples that were visually interpreted. The legends of the GLC maps and the reference datasets were harmonized into 11
general land cover classes. There results show that the map accuracy estimates vary up to 10-16% depending on the homogeneity of
the reference point (HRP) for all the GLC maps. An increase of the HRP resulted in higher overall accuracies but reduced accuracy
confidence for the GLC maps due to less number of accountable samples. The overall accuracy of the SYNMAP was the highest at
any HRP level followed by the GLC2000. The overall accuracies of the maps also varied by up to 10% depending on the definition
of agreement between the reference and map categories in heterogeneous landscape. A careful consideration of heterogeneous
landscape is therefore recommended for future accuracy assessments of land cover maps.
* Corresponding author
1. INTRODUCTION
The consistent and continuous observation of land cover is one
of the most important foundations for understanding the Earth’s
environment and ecosystems (Verburg et al., 2011). Currently,
several global land cover datasets (GLC) have been developed
and these datasets are evolving towards higher spatial resolution
(Gong et al., 2013; Mora et al., 2014) . Most GLC maps were
developed by individual groups as one-time efforts and the
subsequent mapping standards reflect the varied interests,
requirements and methodologies of the originating programs
(Herold et al., 2006). These differences of GLC maps and the
effects of their quality on the model outcome are not always
considered when selecting a map as an input for specific
modeling applications (Verburg et al., 2011). Uncertainties of
GLC maps can result in considerable differences in modeling
outcomes (Hibbard et al., 2010; Nakaegawa, 2011; Verburg et
al., 2011).
The accuracies of GLC maps are assessed using independent
validation datasets and regional maps or cross validated against
training datasets. The results of accuracy assessments of
previous maps indicate that overall area-weighted accuracy is
around 70% for the existing GLC maps (Defourny et al., 2012).
However, the use of different approaches in the GLC map
production (e.g., classification scheme, data sources and
algorithms) as well as in validation data collection (e.g.,
sampling scheme, data source and method of reference
classification) raise inconsistency issues and make map
comparisons difficult. Several comparative analyses of land
cover maps were conducted at regional levels