Trend analysis using non-stationary time series clustering based on the finite element method

Abstract

Abstract. In order to analyze low-frequency variability of climate, it is useful to model the climatic time series with multiple linear trends and locate the times of significant changes. In this paper, we have used non-stationary time series clustering to find change points in the trends. Clustering in a multi-dimensional non-stationary time series is challenging, since the problem is mathematically ill-posed. Clustering based on the finite element method (FEM) is one of the methods that can analyze multidimensional time series. One important attribute of this method is that it is not dependent on any statistical assumption and does not need local stationarity in the time series. In this paper, it is shown how the FEM-clustering method can be used to locate change points in the trend of temperature time series from in situ observations. This method is applied to the temperature time series of North Carolina (NC) and the results represent region-specific climate variability despite higher frequency harmonics in climatic time series. Next, we investigated the relationship between the climatic indices with the clusters/trends detected based on this clustering method. It appears that the natural variability of climate change in NC during 1950-2009 can be explained mostly by AMO and solar activity.

Fractal and Long-Memory Traces in PM10 Time Series in Athens, Greece

“…The scientific research of the topic focuses on understanding the variations of urban air pollutants with respect to time, investigating the contribution of environmental and meteorological parameters to those variations through modeling, forecasting, and predicting future trends and alterations. Noteworthy approaches utilize neural networks [5][6][7][8][9][10][11][12][13][14][15][16], wavelets [17,18], statistical [19,20], and deterministic [21] models, and, currently, cutting-edge techniques based on chaos and complexity (e.g., References [4,[22][23][24][25][26][27][28][29][30][31][32][33][34][35][36]). Since no forecast method is solid, reliable, and accurate enough to sufficiently match all air-contamination time series [37], it is a challenging task to achieve credible estimations.…”

Characterizing the Spatio-Temporal Pattern of Land Surface Temperature through Time Series Clustering: Based on the Latent Pattern and Morphology

“…Specifically, it identifies clusters with homogeneous temporal patterns through the comparison of similarity among the different time series [23]. The research of time series clustering applied to daily temperature pattern recognition is well documented [27][28][29]. All studies utilized the daily temperature data acquired from sparsely distributed in situ stations.…”

An Efficient Method to Improve the Clustering Performance using Hybrid Robust Principal Component Analysis-Spectral biclustering in Rainfall Patterns Identification

“…Cluster analysis has been developed to subdivide observations of similar patterns and dissimilar patterns accordingly into different clusters. Typically, classical clustering techniques in climate data apply the algorithms to either the rows or the columns within time or space domains of the data matrix separately [10][11][12][13][14]. In time-clustering techniques, segments of time are detected where the values of the time series are similar to each other [15].…”