Methods for Detecting Early Warnings of Critical Transitions in Time Series Illustrated Using Simulated Ecological Data

Many dynamical systems, including lakes, organisms, ocean circulation patterns, or financial markets, are now thought to have tipping points where critical transitions to a contrasting state can happen. Because critical transitions can occur unexpectedly and are difficult to manage, there is a need for methods that can be used to identify when a critical transition is approaching. Recent theory shows that we can identify the proximity of a system to a critical transition using a variety of so-called `early warning signals', and successful empirical examples suggest a potential for practical applicability. However, while the range of proposed methods for predicting critical transitions is rapidly expanding, opinions on their practical use differ widely, and there is no comparative study that tests the limitations of the different methods to identify approaching critical transitions using time-series data. Here, we summarize a range of currently available early warning methods and apply them to two simulated time series that are typical of systems undergoing a critical transition. In addition to a methodological guide, our work offers a practical toolbox that may be used in a wide range of fields to help detect early warning signals of critical transitions in time series data.

Combination of similarity measures for time series classification using genetic algorithms

Time series classification deals with the problem of classification of data that is multivariate in nature. This means that one or more of the attributes is in the form of a sequence. The notion of similarity or distance, used in time series data, is significant and affects the accuracy, time, and space complexity of the classification algorithm. There exist numerous similarity measures for time series data, but each of them has its own disadvantages. Instead of relying upon a single similarity measure, our aim is to find the near optimal solution to the classification problem by combining different similarity measures. In this work, we use genetic algorithms to combine the similarity measures so as to get the best performance. The weightage given to different similarity measures evolves over a number of generations so as to get the best combination. We test our approach on a number of benchmark time series datasets and present promising results.

Predictability of nonstationary time series using wavelet and EMD based ARMA models

Research has been undertaken to ascertain the predictability of non-stationary time series using wavelet and Empirical Mode Decomposition (EMD) based time series models. Methods have been developed in the past to decompose a time series into components. Forecasting of these components combined with random component could yield predictions. Using this ideology, wavelet and EMD analyses have been incorporated separately which decomposes a time series into independent orthogonal components with both time and frequency localizations. The component series are fit with specific auto-regressive models to obtain forecasts which are later combined to obtain the actual predictions. Four non-stationary streamflow sites (USGS data resources) of monthly total volumes and two non-stationary gridded rainfall sites (IMD) of monthly total rainfall are considered for the study. The predictability is checked for six and twelve months ahead forecasts across both the methodologies. Based on performance measures, it is observed that wavelet based method has better prediction capabilities over EMD based method despite some of the limitations of time series methods and the manner in which decomposition takes place. Finally, the study concludes that the wavelet based time series algorithm can be used to model events such as droughts with reasonable accuracy. Also, some modifications that can be made in the model have been discussed that could extend the scope of applicability to other areas in the field of hydrology. (C) 2013 Elesvier B.V. All rights reserved.

Fuzzy Classification of Time Series Data

The problem of classification of time series data is an interesting problem in the field of data mining. Even though several algorithms have been proposed for the problem of time series classification we have developed an innovative algorithm which is computationally fast and accurate in several cases when compared with 1NN classifier. In our method we are calculating the fuzzy membership of each test pattern to be classified to each class. We have experimented with 6 benchmark datasets and compared our method with 1NN classifier.

A Comparison of Machine Learning Techniques for Modeling River Flow Time Series: The Case of Upper Cauvery River Basin

Models of river flow time series are essential in efficient management of a river basin. It helps policy makers in developing efficient water utilization strategies to maximize the utility of scarce water resource. Time series analysis has been used extensively for modeling river flow data. The use of machine learning techniques such as support-vector regression and neural network models is gaining increasing popularity. In this paper we compare the performance of these techniques by applying it to a long-term time-series data of the inflows into the Krishnaraja Sagar reservoir (KRS) from three tributaries of the river Cauvery. In this study flow data over a period of 30 years from three different observation points established in upper Cauvery river sub-basin is analyzed to estimate their contribution to KRS. Specifically, ANN model uses a multi-layer feed forward network trained with a back-propagation algorithm and support vector regression with epsilon intensive-loss function is used. Auto-regressive moving average models are also applied to the same data. The performance of different techniques is compared using performance metrics such as root mean squared error (RMSE), correlation, normalized root mean squared error (NRMSE) and Nash-Sutcliffe Efficiency (NSE).

Using long time series of agricultural-derived nitrates for estimating catchment transit times

The estimation of water and solute transit times in catchments is crucial for predicting the response of hydrosystems to external forcings (climatic or anthropogenic). The hydrogeochemical signatures of tracers (either natural or anthropogenic) in streams have been widely used to estimate transit times in catchments as they integrate the various processes at stake. However, most of these tracers are well suited for catchments with mean transit times lower than about 4-5 years. Since the second half of the 20th century, the intensification of agriculture led to a general increase of the nitrogen load in rivers. As nitrate is mainly transported by groundwater in agricultural catchments, this signal can be used to estimate transit times greater than several years, even if nitrate is not a conservative tracer. Conceptual hydrological models can be used to estimate catchment transit times provided their consistency is demonstrated, based on their ability to simulate the stream chemical signatures at various time scales and catchment internal processes such as N storage in groundwater. The objective of this study was to assess if a conceptual lumped model was able to simulate the observed patterns of nitrogen concentration, at various time scales, from seasonal to pluriannual and thus if it was relevant to estimate the nitrogen transit times in headwater catchments. A conceptual lumped model, representing shallow groundwater flow as two parallel linear stores with double porosity, and riparian processes by a constant nitrogen removal function, was applied on two paired agricultural catchments which belong to the Research Observatory ORE AgrHys. The Global Likelihood Uncertainty Estimation (GLUE) approach was used to estimate parameter values and uncertainties. The model performance was assessed on (i) its ability to simulate the contrasted patterns of stream flow and stream nitrate concentrations at seasonal and inter-annual time scales, (ii) its ability to simulate the patterns observed in groundwater at the same temporal scales, and (iii) the consistency of long-term simulations using the calibrated model and the general pattern of the nitrate concentration increase in the region since the beginning of the intensification of agriculture in the 1960s. The simulated nitrate transit times were found more sensitive to climate variability than to parameter uncertainty, and average values were found to be consistent with results from others studies in the same region involving modeling and groundwater dating. This study shows that a simple model can be used to simulate the main dynamics of nitrogen in an intensively polluted catchment and then be used to estimate the transit times of these pollutants in the system which is crucial to guide mitigation plans design and assessment. (C) 2015 Elsevier B.V. All rights reserved.

Reversible jump sampler for autoregressive time series

We use reversible jump Markov chain Monte Carlo (MCMC) methods to address the problem of model order uncertainty in autoregressive (AR) time series within a Bayesian framework. Efficient model jumping is achieved by proposing model space moves from the full conditional density for the AR parameters, which is obtained analytically. This is compared with an alternative method, for which the moves are cheaper to compute, in which proposals are made only for new parameters in each move. Results are presented for both synthetic and audio time series.

Bayesian model selection for time series using Markov chain Monte Carlo

We present a stochastic simulation technique for subset selection in time series models, based on the use of indicator variables with the Gibbs sampler within a hierarchical Bayesian framework. As an example, the method is applied to the selection of subset linear AR models, in which only significant lags are included. Joint sampling of the indicators and parameters is found to speed convergence. We discuss the possibility of model mixing where the model is not well determined by the data, and the extension of the approach to include non-linear model terms.

Predicting chaotic time-series with wavelet networks

A new technique, wavelet network, is introduced to predict chaotic time series. By using this technique, firstly, we make accurate short-term predictions of the time series from chaotic attractors. Secondly, we make accurate predictions of the values and bifurcation structures of the time series from dynamical systems whose parameter values are changing with time. Finally we predict chaotic attractors by making long-term predictions based on remarkably few data points, where the correlation dimensions of predicted attractors are calculated and are found to be almost identical to those of actual attractors.

Direct dynamical test for deterministic chaos and optimal embedding of a chaotic time-series

We propose here a local exponential divergence plot which is capable of providing an alternative means of characterizing a complex time series. The suggested plot defines a time-dependent exponent and a ''plus'' exponent. Based on their changes with the embedding dimension and delay time, a criterion for estimating simultaneously the minimal acceptable embedding dimension, the proper delay time, and the largest Lyapunov exponent has been obtained. When redefining the time-dependent exponent LAMBDA(k) curves on a series of shells, we have found that whether a linear envelope to the LAMBDA(k) curves exists can serve as a direct dynamical method of distinguishing chaos from noise.

Local exponential divergence plot and optimal embedding of a chaotic time-series

we propose here a local exponential divergence plot which is capable of providing a new means of characterizing chaotic time series. The suggested plot defines a time dependent exponent LAMBDA and a ''plus'' exponent LAMBDA+ which serves as a criterion for estimating simultaneously the minimal acceptable embedding dimension, the proper delay time and the largest Lyapunov exponent.