Technical note: A procedure to clean, decompose, and aggregate time series

Abstract

Errors, gaps, and outliers complicate and sometimes invalidate the analysis of time series. While most fields have developed their own strategy to clean the raw data, no generic procedure has been promoted to standardize the pre-processing. This lack of harmonization makes the inter-comparison of studies difficult, and leads to screening methods that can be arbitrary or case-specific. This study provides a generic pre-processing procedure implemented in R (ctbi for cyclic/trend decomposition using bin interpolation) dedicated to univariate time series. Ctbi is based on data binning and decomposes the time series into a long-Term trend and a cyclic component (quantified by a new metric, the Stacked Cycles Index) to finally aggregate the data. Outliers are flagged with an enhanced box plot rule called Logbox that corrects biases due to the sample size and that is adapted to non-Gaussian residuals. Three different Earth science datasets (contaminated with gaps and outliers) are successfully cleaned and aggregated with ctbi. This illustrates the robustness of this procedure that can be valuable to any discipline.