South Australian rainfall - trends and climate drivers

Characterisation of rainfall and identification of potential climate drivers in the South Australia (SA) region are significantly important for the development of climate change adaptation strategies and for its sustainable natural resources management. This study demonstrated the existence of statistically significant trends in SA rainfalls. It also identified the existence of correlations between SA rainfalls and potential climate drivers (SOI, DMI and Niño3.4). Fifty-three high quality rainfall stations distributed across the eight SA natural resources management (NRM) regions have been used for the analyses. Robust statistical techniques for trend identification (Mann-Kendall, Linear Regression and Spearman rho tests) and bivariate correlation analysis (Pearson, Spearman and Kendall tests) have been followed. A bootstrapping method was applied for significance analyses. Increasing annual rainfall trends were observed for the Arid Lands, Alinytjara Wilinara and Adelaide and Mount Lofty Ranges regions whereas decreasing trends were found for the Murray Darling Basin, Eyre Peninsula and South East regions. Both upward and downward trends were identified in the Northern and Yorke region. Seasonally, summer and spring rainfalls exhibited increasing trends in most of the regions whereas autumn and winter rainfalls showed decreasing trends. During Australian summer and autumn, significant correlations with the SOI were observed in the SA south eastern NRM regions. Both SOI and DMI were found significantly correlated with winter rainfalls in the SA south east region. DMI was also found to be correlated during spring in all NRM regions except for the Alinytjara-Wilurara region. Correlation with Niño3.4 was mostly observed in summer in the Arid Lands region. This paper also includes a discussion on possible teleconnection paths for these climate drivers