Acta Geographica Sinica ›› 2019, Vol. 74 ›› Issue (5): 923-934.doi: 10.11821/dlxb201905007

• Climate Change and Surface Processes • Previous Articles     Next Articles

Reliability assessment of global historical forest data in China

YANG Fan1,2,HE Fanneng1(),LI Meijiao1,2,LI Shicheng3   

  1. 1. Key Laboratory of Land Surface Pattern and Simulation, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
    2. University of Chinese Academy of Sciences, Beijing 100049, China
    3. School of Public Administration, China University of Geosciences, Wuhan 430074, China
  • Received:2018-03-07 Revised:2019-01-23 Online:2019-05-25 Published:2019-05-24
  • Contact: HE Fanneng
  • Supported by:
    The Chinese Academy of Sciences Strategic Priority Research Program(XDA19040101);National Natural Science Foundation of China(41671149);The National Key Research and Development Program of China(2017YFA0603304)


Global historical land use datasets play a quite significant role in gaining a profound understanding of the global or regional environmental change. As an important component of these global datasets, the reliability of historical forest data at the regional scale is rarely assessed. Based on the Chinese historical forest dataset (CHFD) that is reestablished by Chinese scholars using historical documents, we conducted an evaluation of the reliability of China's forest data within these global datasets (SAGE, PJ and KK10) adopting comparative analysis from three aspects, including the change tendency of forest area, the area of forestland, and the differences at grid scale. The results indicated that: (1) Although Chinese forest area from multiple datasets was decreasing, there was a large difference in the quantity of forest area. Specifically, the forest area of China from 1700-1990 in SAGE dataset was 20%-40% greater than that in CHFD, while the forest area of China in KK10 dataset from 1700-1850 was approximately 32%-46% greater than that in CHFD. Due to the adoption of regional research results, the total forest area of China within PJ dataset was closen to CHFD, and the quantitative biases of most years were less than 20%. (2) At provincial scale, in terms of the PJ dataset which was relatively close to the CHFD, the proportion of the provinces with large difference in the changing trend was 84%, and the proportion of the provinces with large difference in the quantity could be up to 92%. (3) At grid cell scale, the percentage of grid cells having biases greater than 70% accounted for up to 60%-80%. Therefore, there was an apparent discrepancy of spatiotemporal dynamic patterns between PJ and CHFD datasets. (4) These global datasets failed to reveal the process and pattern characteristics of Chinese forest dynamics in an objective way. The major reasons were that different data sources were used in reconstructing historical forest data within global and regional datasets, and different reconstruction methods at different spatial scales were adopted.

Key words: historical forest, global datasets, reliability assessment, China