Demonstration of a Principal Component Analysis Trajectory Method to Assess Bioremediation Progress at a TCE‐Impacted Site

Abstract

In‐situ bioremediation (ISB) is a popular remediation technology for the treatment of a range of compounds, including chlorinated solvents such as tetrachloroethene and trichloroethene (TCE). Large amounts of data are collected before, during, and after ISB applications to determine amendment approaches, monitor progress and evaluate success. The interpretation of these large datasets can be limited by the tools and techniques used for data analysis, and there is considerable potential in applying data reduction and multivariate techniques used elsewhere to performance monitoring during ISB. In this study, a principal component analysis (PCA) trajectory method was applied to a TCE‐impacted ISB site dataset, as an alternative to the inspection of time series data. The method connected each monitoring well's scores through PCA space to account for temporal changes in multiple analytes across the site. The method was used to separate monitoring well locations into categories that included On‐track and Unsuccessful based on their similarity to background wells in PCA space. The results agreed with those generated using traditional methods (e.g., time series plots) and were able to efficiently summarize large amounts of data to facilitate interpretation. It is expected that this PCA trajectory method could provide a useful screening tool to quickly identify site‐wide trends for the evaluation of ISB performance.

Aircraft Behavior Recognition on Trajectory Data with a Multimodal Approach

“…The data preprocessing stage aims to transform irregular sequences into fixed-size features. Principal component analysis (PCA) [4], dynamic time warping (DTW) [5], and other linear projection methods [6] were adopted to process moving traces. Besides, curve approximation and reconstruction can realize length alignment effectively [7,8].…”