Characterization of forest types in Northeastern China, using multi-temporal SPOT-4 VEGETATION sensor data Xiangming Xiao a, * , Stephen Boles a , Jiyuan Liu b , Dafang Zhuang b , Mingliang Liu b a Institute for the Study of Earth, Oceans and Space, University of New Hampshire, Durham, NH 03824, USA b Institute of Geographical Sciences and Natural Resources, Chinese Academy of Sciences, Beijing 100101, China Received 27 December 2001; received in revised form 15 April 2002; accepted 20 April 2002 Abstract In this study, we explored the potential of multi-temporal SPOT-4 VEGETATION (VGT) sensor data for characterization of temperate and boreal forests in Northeastern China. As the VGT sensor has a short-wave infrared (SWIR) band that is sensitive to vegetation, soil moisture and leaf water content, the Normalized Difference Water Index (NDWI) was calculated in addition to the Normalized Difference Vegetation Index (NDVI). A forest map of Northeast China was generated from an unsupervised classification of 25 10-day VGT composite data (NDVI and NDWI) over the period of March 11 –20, 1999 to November 11 – 20, 1999. Seven different forest categories were distinguished from the 1-km spatial resolution VGT data. The VGT forest map was compared to estimates of forest area derived from Landsat 7 Enhanced Thematic Mapper (ETM+) images. There was a good agreement on spatial distribution and area of forest between the VGT product and the TM product, however, the VGT product provided additional information on forest type. Analysis of NDVI and NDWI over the plant growing season allows for the identification of distinct growth patterns between the different forest types. It is evident that VGT data can be used to provide timely and detailed forest maps with limited ancillary data needed. The VGT-derived forest maps could be very useful as input to biogeochemical models (particularly carbon cycle models) that require timely estimates of forest area and type. D 2002 Published by Elsevier Science Inc. 1. Introduction Northeastern China (Fig. 1) has abundant tree species and a variety of forest types, including evergreen needleleaf forest, deciduous needleleaf forest, deciduous broadleaf forest, and mixed forests (Zheng, Xiao, Guo, & Howard, 2001). Human activities (e.g. forest clear-cutting, selective logging, agricultural encroachment) and natural disturbance (e.g. fire, insects) have resulted in substantial losses of the old-growth forests and fragmentation in forest landscapes (Shao et al., 1996; Chen, 2000; Chen, Zhang, Zhou, & Chen, 2000; Liu, Kondoh, Tateishi, Takamura, & Takeuchi, 2001; Wang, Feng, & Ouyang, 2001). For instance, most of the mixed broadleaf/Korean pine (Pinus koraiensis) forests in Jilin and Liaoning Provinces have been replaced (planta- tion-style) by faster-growing species such as larch (Larix sp.), poplar (Populus sp.) and birch (Betula sp.) (Jiang et al., 1999). Fang, Chen, Peng, Zhao, and Ci (2001) and Fang, Wang, Liu, and Xu (1998) have shown that there is considerable spatial variability in biomass carbon storage and density among forest types in China. For example, the larch forests of Northeast China tend to have higher carbon densities than the oak forests (Fang et al., 1998; Jiang, Peng et al., 1999). Recent carbon cycle studies have indicated that the mid- to high-latitude forests of the Northern Hemi- sphere may serve as a significant carbon sink (Schimel et al., 2001). Timely and accurate information on forest types and areas at the regional scale is needed for natural resource management, carbon cycle studies and modeling of bio- geochemistry, hydrology and climate. Satellite-based remote sensing products provide one option to meet those data needs. A number of earlier studies have used Landsat Thematic Mapper (TM) images to document forest types and changes at Changbai Mountain, Jilin Province, where an International Biosphere Reserve was established in 1979 (Shao et al., 1996; Zheng, Wallin, & Hao, 1997; Liu, Kondoh et al., 2001). Because of frequent cloud cover and the long re-visit time (16 days) of Landsat, it is difficult to acquire cloud-free images for monitoring the changes in forest types at short-term intervals (e.g. yearly). Over the last two decades, numerous studies of large-scale mapping of land cover and land use have explored data 0034-4257/02/$ - see front matter D 2002 Published by Elsevier Science Inc. PII:S0034-4257(02)00051-2 * Corresponding author. Tel.: +1-603-862-3818; fax: +1-603-862-0188. E-mail address: xiangming.xiao@unh.edu (X. Xiao). www.elsevier.com/locate/rse Remote Sensing of Environment 82 (2002) 335 – 348