PP41D-1422:
Towards Better Calibration of Modern Palynological Data against Climate: A Case Study in Osaka Bay, Japan

Thursday, 18 December 2014
Ikuko Kitaba, Kobe University, Kobe, Japan, Takeshi Nakagawa, Ritsumeikan University, Kusatsu, Japan; Newcastle University, Newcastle Upon Tyne, United Kingdom, Erin McClymont, University of Durham, Durham, DH1, United Kingdom, David L Dettman, University of Arizona, Tucson, AZ, United States, Keitaro Yamada, Kyoto University, Kyoto, Japan, Keiji Takemura, Kyoto Univ, Beppu, Japan and Masayuki Hyodo, Kobe Univ, Kobe, Japan
Abstract:
Many of the difficulties in the pollen fossil-based paleoclimate reconstruction in coastal regions derive from the complex sedimentary processes of the near-shore environment. In order to examine this problem, we carried out pollen analysis of surface sediments collected from 35 sites in Osaka Bay, Japan. Using the biomisation method, the surrounding vegetation was accurately reconstructed at all sites. Applying the modern analogue technique to the same data, however, led to reconstructed temperatures that were lower by ca. 5 deg. C and precipitation amounts higher by ca. 5000 mm than the current sea level climate of the region. The range of reconstructed values was larger than the reconstruction error associated with the method. The principal component analysis shows that the surface pollen variation in Osaka Bay reflects sedimentary processes. This significant error associated with the quantitative climatic reconstruction using pollen data is attributed to the fact that the pollen assemblage is not determined solely by climate but reflects non-climatic influences. The accuracy and precision of climatic reconstruction can be improved significantly by expanding counts of minor taxa. Given this result, we re-examined the reconstructed climate using Osaka Bay palynological record reported in Kitaba et al. (2013). This new method did not significantly alter the overall variation in the reconstructed climate, and thus we conclude that the reconstruction was generally reliable. However, some intervals were strongly affected by depositional environmental change. In these, a climate signal can be extracted by excluding the patterns that arise from coastal sedimentation.