A0447
Title: The North Atlantic Oscillation and monthly precipitation in selected European locations: A time series approach
Authors: Timo Terasvirta - Aarhus University (Denmark) [presenting]
Annastiina Silvennoinen - Queensland University of Technology (Australia)
Jian Kang - Dongbei University of Finance and Ecomnomics (China)
Changli He - Tianjin University of Finance and Economics (China)
Abstract: The relationship between the monthly precipitation in 30 European cities and towns, and two Algerian ones, and the North Atlantic Oscillation (NAO) index is characterised using the vector seasonal shifting mean and covariance autoregressive model, extended to contain exogenous variables. Central statistical and time series features of the model are considered before moving on to discussing data and the empirical results. The results, based on the monthly time series from 1851 up until 2022, include shifting monthly means for the rainfall series and, even more importantly, the estimated coefficients of the exogenous NAO variable. The empirical results, based on long monthly time series, agree with previous ones in the literature in that the NAO has its strongest effect on precipitation during the winter months. The negative effect is particularly strong in Western Europe and in the Mediterranean rim. The effect in northern locations is positive for the winter months and as such opposite to the corresponding effect in the west and the Mediterranean. In Western and Central Europe, the effect of the NAO is generally negative but not as strong as in the Mediterranean. There is plenty of individual variation, however. The model also contains a time-varying error covariance matrix that is decomposed into time-varying variances and correlations. The constancy of the error variances is tested and the results are reported.
