Reanalysis of Principal Component Analysis to Resume Past Vegetation and Environment

Zhang Jiahua

DOI: https://doi.org/10.1007/s11769-997-0015-z

IF: 3.101

1998-01-01

Chinese Geographical Science

Abstract:In the paper 11 pollen types were selected from the original pollen and spore records of 54 samples in Fenzhuang profile of Beijing region, then 54 × 11 matrix was set up. Based on the pollen data of Fenzhuang profile from the Late Pleistocene to the Early Holocene, the paper gives a further analysis to subdivide zone of the pollen assemblages and changes of paleovegetation and paleoenvironment in this area using principal component analysis (PCA). PCA can play an complementary role for subdividing zone, at the same time, it can eliminate the difference to subdivide zone of the pollen assemblages caused by artificial factors. Finally to the biodiversity in plant communities and complex geographical environment in Beijing region, using great dispersal of accumulative percent of total in PCA (for cumulative rate only 57% from the first three factors of PCA), may sufficiently correspond to the environmental characteristic, so as to recover past climate condition much better.