Use of principal components analysis (PCA) on estuarine sediment datasets: The effect of data pre-treatment |
| |
| Authors: | MK Reid KL Spencer |
| |
| Institution: | 1. Institute of Inorganic Chemistry AS CR, v.v.i., ?e?, Czech Republic;2. Faculty of Environment, J.E. Purkyně University in Ústí n.L., Czech Republic;3. Faculty of Science, Palacký University, Olomouc, Czech Republic;4. Faculty of Science, Charles University, Prague, Czech Republic;5. Department of Physical Geography and Geoinformatics, University Szeged, Hungary;1. Key Laboratory of Submarine Geosciences and Technology, MOE, Ocean University of China, Qingdao 266100, China;2. Laboratory for Marine Geology, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266061, China;1. Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, 1799 Jimei Boulevard, Xiamen City 361021, China;2. Key Laboratory of Tropical Marine Environmental Dynamics, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou 510301, China;3. School of Environment, Northeast Normal University, Changchun 130117, China;1. State Key Laboratory of Environmental Aquatic Chemistry, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, P.O. Box 2871, Beijing 100085, China;2. Beijing Climate Change Respond Research and Education Centre, Beijing University of Civil Engineering and Architecture, 1 Zhanlan Rd., Xicheng, Beijing 100044, China;3. Department of Chemistry, University of Oslo, 0315 Oslo, 1033, Norway;4. Department of Biosciences, University of Oslo, 0316 Oslo, 1066, Norway;5. Miyun Reservoir Administration, Xiwenzhuang, Miyun, Beijing 101512, China |
| |
| Abstract: | Principal components analysis (PCA) is a multivariate statistical technique capable of discerning patterns in large environmental datasets. Although widely used, there is disparity in the literature with respect to data pre-treatment prior to PCA. This research examines the influence of commonly reported data pre-treatment methods on PCA outputs, and hence data interpretation, using a typical environmental dataset comprising sediment geochemical data from an estuary in SE England. This study demonstrated that applying the routinely used log (x + 1) transformation skewed the data and masked important trends. Removing outlying samples and correcting for the influence of grain size had the most significant effect on PCA outputs and data interpretation. Reducing the influence of grain size using granulometric normalisation meant that other factors affecting metal variability, including mineralogy, anthropogenic sources and distance along the salinity transect could be identified and interpreted more clearly. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
