Time trend estimation with breaks in temperature time series
This paper deals with the modelling of the global and northern and southern hemispheric anomaly temperature time series using a novel technique based on segmented trends and fractional integration. We use a procedure that permits us to estimate linear time trends and orders of integration at various subsamples, where the periods for the changing trends are endogenously determined by the model. Moreover, we use a non-parametric approach (Bloomfield, 1973) for modelling the I(0) deviation term. The results show that the three series (global, northern and southern temperatures) can be well described in terms of fractional integration with the orders of integration around 0.5 in the three cases. The coefficients associated to the time trends are statistically significant in all subsamples for the three series, especially during the final part of the sample, giving then some support to the global warming theories.
Year of publication: |
2008-11-20
|
---|---|
Authors: | Gil-Alana, Luis A. |
Institutions: | School of Economics and Business Administration, University of Navarra |
Saved in:
freely available
Saved in favorites
Similar items by person
-
Fractional Integration and Cointegration in US Financial Time Series Data
Gil-Alana, Luis A., (2012)
-
An Analysis of Oil Production by OPEC Countries: Persistence, Breaks, and Outliers
Barros, Carlos Pestana, (2011)
-
US Disposable Personal Income and Housing Price Index: A Fractional Integration Analysis
Caporale, Guglielmo Maria, (2011)
- More ...