M.Sc. Tezi Görüntüleme | |||||||||||||||||||||
|
|
||||||||||||||||||||
Summary: Nowadays, time series analysis make possible to examine the changes occurring in many data groups. In this study, coordinate time series were analyzed at IGS (International GNSS Service) stations, which are stationary to various parts of the world but are trend and seasonal in time due to effects such as crustal movements and atmospheric loads. In the IGS coordinate series based on GNSS (Global Navigation Satellite Systems) systems, ANKR, ISTA and TUBI stations in Turkey, were used for the analysis of trend and seasonal oscillations. In the coordinate series, firstly outliers measurements were determined by k-sigma method. After examining the data cuttings in the series, outliers measurements were extracted from the data. In order to examine the existence of trends and seasonality in the series, autocorrelation analysis was applied. For trend analysis, Linear Regression Model, Simple Moving Average and EMD (Empricial Mode Decomposition) method based on the HHT (Hilbert Huang Transformation) was applied. The HT (Hilbert Transformation) based on HHT was applied for the IMF (Intrinsic Mode Functions) obtained from the EMD method and instantaneous frequency changes in the series have been examined.
Key Words: IGS Stations, Coordinate Time Series, Trend, Seasonality, Autocorrelation, Linear Regression, Moving Average, Hilbert Huang Transform, Empricial Mode Decomposition, Hilbert Transformation. |