A simultaneous analysis of gradual and abrupt changes in Canadian low streamflows

Ehsanzadeh, Eghbal; Ouarda, Taha B. M. J.; Saley, Hadiza Moussa

doi:10.1002/hyp.7861

Cited by 47 publications

(24 citation statements)

References 37 publications

(55 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Bayesian approaches have been developed for climate or hydrological applications by Seidou & Ouarda [54] and Hannart & Naveau [55]. Techniques able to detect several shifts in the mean or in a linear regression model have been useful to study shifts in temperature, precipitation and climate indices such as the PDO [1,10] and also useful to detect several shifts in streamflows [54,56].…”

Section: (C) Several Shiftsmentioning

confidence: 99%

Change-point analysis as a tool to detect abrupt climate variations

Beaulieu

Chen

Sarmiento

2012

Phil. Trans. R. Soc. A.

146

View full text Add to dashboard Cite

Recently, there have been an increasing number of studies using change-point methods to detect artificial or natural discontinuities and regime shifts in climate. However, a major drawback with most of the currently used change-point methods is the lack of flexibility (able to detect one specific type of shift under the assumption that the residuals are independent). As temporal variations in climate are complex, it may be difficult to identify change points with very simple models. Moreover, climate time series are known to exhibit autocorrelation, which corresponds to a model misspecification if not taken into account and can lead to the detection of non-existent shifts. In this study, we extend a method known as the informational approach for change-point detection to take into account the presence of autocorrelation in the model. The usefulness and flexibility of this approach are demonstrated through applications. Furthermore, it is highly desirable to develop techniques that can detect shifts soon after they occur for climate monitoring. To address this, we also carried out a simulation study in order to investigate the number of years after which an abrupt shift is detectable. We use two decision rules in order to decide whether a shift is detected or not, which represents a trade-off between increasing our chances of detecting a shift and reducing the risk of detecting a shift while in reality there is none. We show that, as of now, we have good chances to detect an abrupt shift with a magnitude that is larger than that of the standard deviation in the series of observations. For shifts with a very large magnitude (three times the standard deviation), our simulation study shows that after only 4 years the probabilities of shift detection reach nearly 100 per cent. This reveals that the approach has potential for climate monitoring.

show abstract

Section: (C) Several Shiftsmentioning

confidence: 99%

Change-point analysis as a tool to detect abrupt climate variations

Beaulieu

Chen

Sarmiento

2012

Phil. Trans. R. Soc. A.

146

View full text Add to dashboard Cite

show abstract

“…A number of detected change points with the highest probability of occurrence are then chosen. Then, Bayesian inference provides the time position of each selected change point and its probability distribution of occurrence (for more details see Ehsansazeh, Ouarda, & Saley, 2011). In this study, we apply this method to find any possible change points in the SOI time series in order to compare the statistical and stochastic features of SOI before and after any detected change point.…”

Section: Methodology a Garch Modelling Of Soi Time Seriesmentioning

confidence: 99%

Testing and Modelling the Volatility Change in ENSO

Modarres¹,

Ouarda²

2013

Atmosphere-Ocean

View full text Add to dashboard Cite

The El Niño-Southern Oscillation (ENSO) is by far the most energetic climate signal. Any change in ENSO characteristics will have serious consequences for the global climate system. This work suggests a different view at the change in ENSO volatility in addition to change in its descriptive statistics. The volatility or the conditional variance of ENSO is tested and modelled using both the Autoregressive Moving Average-Generalized Autoregressive Conditional Heteroscedasticity (ARMA-GARCH) error model and the GARCH model, to investigate the change in the short-run and long-run persistency of the second-order moment of ENSO before and after a change point detected by a Bayesian change point analysis. Nonparametric tests revealed a significant change in descriptive statistical characteristics such as the mean, the (unconditional) variance, and the probability distribution of ENSO after a change point in 1975. An Engle's test did not show heteroscedasticity in the random process (residuals) of the Southern Oscillation Index (SOI) time series before 1975 although heteroscedasticity increased and appeared after 1975. The GARCH model indicates an increasing short-run persistency after 1975 and decreasing long-run persistency. A seasonal shift in extreme heteroscedasticity is observed from summer to winter. In addition, the non-linearity and nonstationarity of the SOI volatility have increased in recent decades. This may be caused by an increase in frequency and magnitude of extreme volatilities after 1975. The results of this study indicate that ENSO has become more dynamic and uncertain in recent decades. The increase in the frequency of extreme events together with extreme conditional variance after 1975 may increase the prediction uncertainty of ENSO-driven climate phenomena.RÉSUMÉ [Traduit par la rédaction] Le El Niño-oscillation australe (ENSO) est de loin le signal climatique le plus énergétique. Tout changement dans les caractéristiques de l'ENSO aura de sérieuses conséquences sur le système climatique planétaire. Le présent travail propose un regard différent sur le changement dans la volatilité de l'ENSO en plus du changement dans ses statistiques descriptives. Nous testons et modélisons la volatilité ou variance conditionnelle de l'ENSO en nous servant à la fois du modèle d'erreur ARMA-GARCH (moyenne mobile autorégressive -hétéroscédasticité conditionnelle autorégressive généralisée) et du modèle GARCH pour étudier le changement à court terme et à long terme de la persistance du moment de second ordre de l'ENSO avant et après un point de changement détecté par une analyse bayésienne de point de changement. Des tests non paramétriques ont révélé un changement significatif dans les caractéristiques statistiques descriptives comme la moyenne, la variance (inconditionnelle) et la distribution de probabilités de l'ENSO après un point de changement en 1975. Un test d'Engle n'a pas montré d'hétéroscédasticité dans le processus aléatoire (résidus) de la série chronologique de l'indice d'oscillation australe ...

show abstract

“…It may lead to inaccurate interpretations of the MK test, as a time series exhibiting positive autocorrelation causes the effective sample size to be less than the actual sample size, thereby increasing the variance and the possibility of detecting significant trends when in fact, there are no trends (Hamed and Rao 1998;Ehsanzadeh et al 2011). On contrary, the existence of negative autocorrelation in a time series enhances the possibility of accepting the null hypothesis (absence of significant trends), when actually, there are significant trends (Ehsanzadeh et al 2011). To take care of this issue Hamed and Rao (1998) proposed a modified approach, in which the calculation of the variance of the test statistics S is altered as given by an empirical formula.…”

Section: Modified Mann-kendall (Mmk) Testmentioning

confidence: 99%

Trend analysis of sediment flux time series from tropical river basins in India using non-parametric tests and multiscale decomposition

Adarsh¹,

VishnuPriya²,

Narayanan³

et al. 2016

Model. Earth Syst. Environ.

View full text Add to dashboard Cite

This paper first analyzed the trend of annual sediment flux time series from seven major basin outlet stations in India using the non-parametric Mann-Kendall (MK) method. Then the non-linear trends in these series were extracted using the empirical mode decomposition (EMD) method. Similar procedure is followed upon annual rainfall of the basin and a comparison of trend is made with that of sediment load. Results of trend analysis study showed a reduction in sediment flux in five out of seven tropical river basins in India despite the increase in rainfall in three among them. Despite the significant reduction in rainfall, the sediment flux from Chennur station in Pennar basin showed an increasing trend, which is attributed to large flow diversion in the upper reaches of the station. The application of continuous wavelet transform (CWT) analysis confirmed that anthropogenic impact is the major driver behind the sediment flux variability of tropical river basins in India. Then for a microscopic examination of trend in monthly sediment flux data from different stations, first the change points in them were identified by the Hubert's segmentation procedure. For each of the identified segments, the trend is examined using the MK test and EMD method. The nature of trend captured by both methods is similar for most of the cases with discrepancies for few cases. The discrete wavelet transform (DWT) analysis is applied for such segments with discrepancies following extensive sensitivity analysis on mother wavelet type and decomposition levels. The similar nature of trend by DWT as that of EMD confirmed that EMD method is successful in capturing the 'true' inherent non-linear trend in sediment flux series.

show abstract

A simultaneous analysis of gradual and abrupt changes in Canadian low streamflows

Cited by 47 publications

References 37 publications

Change-point analysis as a tool to detect abrupt climate variations

Change-point analysis as a tool to detect abrupt climate variations

Testing and Modelling the Volatility Change in ENSO

Trend analysis of sediment flux time series from tropical river basins in India using non-parametric tests and multiscale decomposition

Contact Info

Product

Resources

About