Ceylan Yozgatligil

Comparison of missing value imputation methods in time series: the case of Turkish meteorological data

This study aims to compare several imputation methods to complete the missing values of spatio–temporal meteorological time series. To this end, six imputation methods are assessed with respect to various criteria including accuracy, robustness, precision, and efficiency for artificially created missing data in monthly total precipitation and mean temperature series obtained from the Turkish State Meteorological Service. Of these methods, simple arithmetic average, normal ratio (NR), and NR weighted with correlations comprise the simple ones, whereas multilayer perceptron type neural network and multiple imputation strategy adopted by Monte Carlo Markov Chain based on expectation–maximization (EM-MCMC) are computationally intensive ones. In addition, we propose a modification on the EM-MCMC method. Besides using a conventional accuracy measure based on squared errors, we also suggest the correlation dimension (CD) technique of nonlinear dynamic time series analysis which takes spatio–temporal dependencies into account for evaluating imputation performances. Depending on the detailed graphical and quantitative analysis, it can be said that although computational methods, particularly EM-MCMC method, are computationally inefficient, they seem favorable for imputation of meteorological time series with respect to different missingness periods considering both measures and both series studied. To conclude, using the EM-MCMC algorithm for imputing missing values before conducting any statistical analyses of meteorological data will definitely decrease the amount of uncertainty and give more robust results. Moreover, the CD measure can be suggested for the performance evaluation of missing data imputation particularly with computational methods since it gives more precise results in meteorological time series.

Clustering current climate regions of Turkey by using a multivariate statistical method

In this study, the hierarchical clustering technique, called Ward method, was applied for grouping common features of air temperature series, precipitation total and relative humidity series of 244 stations in Turkey. Results of clustering exhibited the impact of physical geographical features of Turkey, such as topography, orography, land–sea distribution and the high Anatolian peninsula on the geographical variability. Based on the monthly series of nine climatological observations recorded for the period of 1970–2010, 12 and 14 clusters of climate zones are determined. However, from the comparative analyses, it is decided that 14 clusters represent the climate of Turkey more realistically. These clusters are named as (1) Dry Summer Subtropical Semihumid Coastal Aegean Region; (2) Dry-Subhumid Mid-Western Anatolia Region; (3 and 4) Dry Summer Subtropical Humid Coastal Mediterranean region [(3) West coast Mediterranean and (4) Eastern Mediterranean sub-regions]; (5) Semihumid Eastern Marmara Transition Sub-region; (6) Dry Summer Subtropical Semihumid/Semiarid Continental Mediterranean region; (7) Semihumid Cold Continental Eastern Anatolia region; (8) Dry-subhumid/Semiarid Continental Central Anatolia Region; (9 and 10) Mid-latitude Humid Temperate Coastal Black Sea Region [(9) West Coast Black Sea and (10) East Coast Black Sea sub-regions]; (11) Semihumid Western Marmara Transition Sub-region; (12) Semihumid Continental Central to Eastern Anatolia Sub-region; (13) Rainy Summer Semihumid Cold Continental Northeastern Anatolia Sub-region; and (14) Semihumid Continental Mediterranean to Eastern Anatolia Transition Sub-region. We believe that this study can be considered as a reference for the other climate-related researches of Turkey, and can be useful for the detection of Turkish climate regions, which are obtained by a long-term time course dataset having many meteorological variables.

Impact of alternating wet and dry periods on long-term seasonal phosphorus and nitrogen budgets of two shallow Mediterranean lakes

The water balance, with large seasonal and annual water level fluctuations, has a critical influence on the nitrogen and phosphorus dynamics of shallow lakes in the semi-arid climate zone. We constructed seasonal water and nutrient budgets for two connected shallow lakes, Lakes Mogan and Eymir, located in Central Anatolia, Turkey. The study period covered 20years with alternations between dry and wet years as well as restoration efforts including sewage effluent diversion and biomanipulations in Lake Eymir. Both lakes experienced a 1-2m water level drop during a drought period and a subsequent increase during the wet period, with seasonal water level fluctuations of 0.60 to 0.70m. During wet years with high water levels, small seasonal differences were observed with a nutrient peak in spring caused by external loading and nutrient loss via retention during summer. During years with low water levels, nutrient concentrations increased due to internal and external loading, exacerbated by evaporative water loss. In Lake Eymir, a shift to eutrophic conditions with turbid water occurred under low water level conditions and consequent internal loading of P from the sediment, causing high nutrient concentrations in summer. Our results indicate a threat of lakes drying out in the semi-arid climate zone if evaporation increases and precipitation decreases as anticipated from the global climate change predictions. In addition, our results show the influence of the water balance on the eutrophication of shallow lakes in the Mediterranean climate zone and highlight the ultimate consequences for lake management.

Determining the Climate Zones of Turkey by Center-Based Clustering Methods

There is a growing evidence that the climate change has already had significant impacts on the world’s physical, biological, and human systems, and it is expected that these impacts will become more severe in the near future. Alterations in the weather patterns and the existence of extreme events can be considered as important indicators of this change. The validity of this reality can be judged by analyzing climate data thoroughly. In this study, for determining the climate zones of Turkey, temperature measures obtained from the Turkish State Meteorological Service stations in the period 1950–2006 are examined by using two center-based clustering methods, namely k-means and fuzzy k-means. The clusters obtained from these methods are compared using objective criteria. They are also evaluated subjectively by the domain experts.

An analysis of the prevalence of malaria in Turkey over the last 85 years

Abstract Background: Affecting 106 countries, malaria is a major global burden. Though intensive antimalaria efforts in Turkey have been successful in bringing down the number of cases, historically malaria was a serious public health concern. Methods: This paper reviews the prevalence rates of malaria in Turkey over the last 85 years (1925–2010). The time series of malaria prevalence was evaluated for possible structural changes by using Chow breakpoint tests and regression models using dummy variables, with autocorrelated errors and generalized autoregressive conditional heteroscedasticity models to assess the impact of volatility in prevalence. Results: Seventy-eight cases of malaria were diagnosed in Turkey in 2010. Malaria prevalence rates in the country show a statistically significant volatility, which underlines the fragility of efforts to control the disease. Conclusions: It is necessary to analyse the national malaria control programme to evaluate to what extent its programmatic capacity, financial resources, and political commitment are sufficient to avoid eroding the gains that have already been made and, ultimately, eradicate malaria. It is essential that there should be no lessening in the long-standing efforts to reduce malaria.

Representation of Multiplicative Seasonal Vector Autoregressive Moving Average Models

Time series often contain observations of several variables and multivariate time series models are used to represent the relationship between these variables. There are many studies on vector autoregressive moving average (VARMA) models, but the representation of multiplicative seasonal VARMA models has not been seriously studied. In a multiplicative vector model, such as a seasonal VARMA model, the representation is not unique because of the noncommutative property of matrix multiplication. In this article, we carefully examine the consequences of different model representations on parameter estimation and forecasting through numerical illustrations, simulation, and the analysis of a housing starts and housing sales dataset.

Modified Maximum-Likelihood Method for Non-Normal Time Series Revisited

Abstract The modified maximum-likelihood method has recently been applied to some non-normal time series models. Our evaluation of these applications revealed that several of the information matrices given in these studies are not correct due to incorrect evaluation of the process mean, and that the estimators for some of the models with a location parameter are not correct. We correct these results. We address to several other issues and propose modifications. We also made some additional simulations, especially for the location parameter case for which there were a very limited number of previous results. Our results indicate that estimations with a location parameter are not as successful as those with no location parameter, and also that the convergence properties of the method are not very favourable.

Temporal clustering of time series via threshold autoregressive models: application to commodity prices

The primary aim in this study is grouping time series according to the similarity between their data generating mechanisms (DGMs) rather than comparing pattern similarities in the time series trajectories. The approximation to the DGM of each series is accomplished by fitting the linear autoregressive and the non-linear threshold autoregressive models, and outputs of the estimates are used for feature extraction. Threshold autoregressive models are recognized for their ability to represent nonlinear features in time series, such as abrupt changes, time-irreversibility and regime-shifting behavior. The proposed clustering approach is mainly based on feature vectors derived from above-mentioned models estimates. Through the use of the proposed approach, one can determine and monitor the set of co-moving time series variables across the time. The efficiency of the proposed approach is demonstrated through a simulation study and the results are compared with other proposed time series clustering methods. An illustration of the proposed clustering approach is given by application to several commodity prices. It is expected that the process of determining the commodity groups that are time-dependent will advance the current knowledge about temporal behavior and the dynamics of co-moving and coherent prices, and can serve as a basis for multivariate time series analyses. Furthermore, generating a time varying commodity prices index and sub-indexes can become possible. Findings suggested that clusters of the prices series have been affected with the global financial crisis in 2008 and the data generating mechanisms of prices and so the clusters of prices might not be the same across the entire time-period of the analysis.

An analysis of the incidence of measles in Turkey since 1960

BACKGROUND/AIM The aims of this study were to evaluate measles incidence and the effect of elimination strategy interventions on the disease from 1960 to 2014 in Turkey. The administration of measles vaccine started in the rural regions in 1970; it was carried out as a campaign along with the National Vaccine Campaign in 1985, and it has been employed as combined measles, mumps, and rubella under the scope of the Measles Elimination Program (MEP) since 2006 in Turkey. While a dramatic decrease in the reporting of measles was observed between 2000 and 2010, the number of the cases has increased since 2011. MATERIALS AND METHODS The time series of measles incidence was evaluated for possible structural changes with regression models using dummy variables, autocorrelated with error terms. RESULTS The incidence of measles showed a statistically significant decline between 1985 and 1988 (P = 0.0072) and between 2005 and 2011 (P < 0.0001). However, a statistically significant increase in incidence was noted after 2013 (P = 0.0008). CONCLUSION Over the last 54 years, the pattern of measles cases demonstrated a significant decline in incidence. However, the increase in incidence in 2013 should be carefully analyzed and interpreted in terms of the MEP.