Bright Emmanuel Owusu and Nittaya McNeil1
Pertanika Journal of Tropical Agricultural Science, Volume 26, Issue 2, April 2018
Keywords: Australia, factor analysis, gamma generalised linear models, multiple regression, rainfall
Published on: 30 Apr 2018
The present study uses statistical methods, specifically multiple regression (MR) and gamma generalised linear models (Gamma GLM), for the modelling and analysis of period rainfall variability in Australia during the period 1950-2013. The data from 92 observational stations were collected from the Australian Bureau of Meteorology. Factor analysis was first used to group the 92 stations into eight rainfall regions. The consecutive five-day mean rainfall in a year (period) in each region is then modelled using MR and the Gamma GLM. The results show that the models fit the data quite well in all the regions, but the MR model did better than the Gamma GLM in some of the regions. The MR models revealed three rainfall groupings, and each group had diverse rainfall patterns and trends. Significant decreasing annual rainfall trend was revealed in the southwest and the north southeast regions. In contrast, significant increasing annual rainfall trends were found in the north and the northwest regions.
ISSN 1511-3701
e-ISSN 2231-8542