AN INNOVATION STATE SPACE APPROACH FOR TIME SERIES FORECASTING

1993 ◽  
Vol 14 (6) ◽  
pp. 589-601 ◽  
Author(s):  
Gaëtan Libert ◽  
Liang Wang ◽  
Bao Liu
Keyword(s):  
Time Series ◽  
State Space ◽  
Time Series Forecasting ◽  
State Space Approach ◽  
Space Approach

scholarly journals Forecasting compositional time series: A state space approach

2017 ◽  
Vol 33 (2) ◽  
pp. 502-512 ◽  
Author(s):  
Ralph D. Snyder ◽  
J. Keith Ord ◽  
Anne B. Koehler ◽  
Keith R. McLaren ◽  
Adrian N. Beaumont
Keyword(s):  
Time Series ◽  
State Space ◽  
State Space Approach ◽  
Space Approach

scholarly journals Analyzing time varying trends in stratospheric ozone time series using state space approach

2013 ◽  
Vol 13 (8) ◽  
pp. 20503-20530 ◽  
Author(s):  
M. Laine ◽  
N. Latva-Pukkila ◽  
E. Kyrölä
Keyword(s):  
Time Series ◽  
Statistical Model ◽  
State Space ◽  
Stratospheric Ozone ◽  
Original Data ◽  
Seasonal Effect ◽  
Practical Implementation ◽  
State Space Approach ◽  
Uncertainty Estimates ◽  
Space Approach

Abstract. We describe a hierarchical statistical state space model for ozone profile time series. The time series are from satellite measurements by the SAGE II and GOMOS instruments spanning years 1984–2011. The original data sets are combined and gridded monthly using 10° latitude bands, and covering 25–55 km with 1 km vertical spacing. In the analysis, mean densities are studied separately for 25–35 km, 35–45 km, and 45–55 km layers, also. Model components include level, trend and seasonal effect with solar activity and Quasi-Biennial Oscillations as proxy variables. We will show how the chosen statistical model is well suited for trend analysis of atmospheric time series that are not stationary but can exhibit both slowly varying and abrupt changes in the distributional properties. The dynamic linear model state space approach provides well defined statistical model for assessing the long term background changes in the ozone time series. The modelling assumptions can be evaluated and the method provides realistic uncertainty estimates for the model based statements on the quantities of interest. We discuss the methodological challenges and practical implementation. The modelling result agree with the hypothesized trend change point for stratospheric ozone at around the year 1997 for mid latitude regions. This is a companion article to Kyrölä et al. (2013).

Sign in / Sign up

Export Citation Format

Share Document