loading page

Systematic Identification of the Drivers of Regional Hydroclimate Variability in Southern Hemispheric Wet Seasons using a Causal Discovery Method
  • Woon Mi Kim,
  • Christian L.E Franzke,
  • Christoph C Raible
Woon Mi Kim
Climate and Environmental Physics, University of Bern

Corresponding Author:woonmi.kim@climate.unibe.ch

Author Profile
Christian L.E Franzke
Center for Climate Physics, Institute for Basic Science
Author Profile
Christoph C Raible
Climate and Environmental Physics, University of Bern
Author Profile


We analyze the drivers of hydroclimate variability in northeast Brazil and southeast Australia using a causal discovery algorithm (CDA). The CDA-based regression models are also generated to assess the predictive skills of this approach. Hydroclimate variability is quantified using the standardized precipitation index and standardized precipitation evapotranspiration index with different time-scales. The CDA highlights the influences of the tropical Atlantic and tropical Pacific on droughts in northeast Brazil. Droughts in southeast Australia are connected to the tropical Pacific and Southern Annual mode. For drought indices with long time-scales, the indices themselves are involved as drivers of hydroclimate conditions reflecting the persistence of long-lasting wet or dry events. The performance of the CDA-based prediction models increases with the time-scales of drought indices and decreases with the lead times of prediction. Overall, our analysis shows the usefulness of CDA to identify causal links in the regional hydroclimate in the southern Hemisphere.