Regression Models with STARMA Errors: An Application to the Study of Temperature Variations in the Antarctic Peninsula
Motivated by spatio-temporal problems that occur in many areas such as environment, geography etc., we propose multivariate regression models with space-time ARMA errors. The methods of model identification and estimation are studied. The asymptotic properties of the estimators have been derived and simulations are provided. The methodology is applied to the analysis of monthly mean surface temperatures at five locations in the Antarctic Peninsula. This area of Antarctic is of great concern to climatologists, as it is believed that there is higher rate of warming compared to the rest. During the period from January 1978 to December 1998, the temperatures at all the five locations in the Antarctic Peninsula have increased. Substantial warming were detected at Faraday/Vernadsky and Rothera stations with the warming rate of 1.07 ∘ C/decade and 1.08 ∘ C/decade respectively and both trends are significant at level 1%.
KeywordsRegression Model Weighting Matrix Linear Regression Model Antarctic Peninsula Multivariate Regression Model
Unable to display preview. Download preview PDF.
- 4.Folland C, Karl T, Christy J, Clarke R, Gruza G, Jouzel J, Mann ME, Oerlemans J, Salinger M, Wang S-W, et al (2001) Observed climate variability and change. In: Houghton JT, Ding Y, Griggs DJ, Noguer M, van der Linden PJ, Dai X, Maskell K, Johnson CA (eds) Climate change 2001: The scientific basis. Cambridge University Press, Cambridge, pp 99–182Google Scholar
- 10.Subba Rao T, Antunes AM (2003) Spatio-temporal modelling of temperature time series: a comprative study. In: Schonberg F, Brillinger DR, Robinson E (eds) Time series analysis and applications to geophysical systems, IMA Publications, Springer, New York, 123–150Google Scholar