Causes and Predictability of Multi-year Variability of Northern Australia Wet-season Rainfall (#202)
The rainfall over Northern Australia (NA, north of 26°S) during the northern wet season (October to April) shows a large variation on seasonal-to-multiyear timescales, which has profound impacts on many agricultural sectors. Beyond seasonal timescales, however, the underlying source of multi-year wet/dry conditions over the NA is still unclear. Here, we investigate the nature of multi-year rainfall variability and its potential predictability during 1900-present. The multi-year rainfall variations are found to exhibit coherence across the western and eastern halves of the NA. Across the east, wet conditions are associated with a La Niña-like sea-surface temperature (SST) pattern in the Pacific, while dry conditions are associated with an El Niño Modoki-like SST pattern. These SST patterns are meridionally broader than those typically associated with a La Niña or El Niño event and bear a strong resemblance to Inter-decadal Pacific Oscillations (IPO). Across the west, wet conditions show similar association with a La Niña-like SST pattern, while dry conditions show a very weak association with tropical SST. This implies limited predictability of multi-year episodes of dry conditions in the west, whereas wet conditions are predictable to the degree that the cold phase of the IPO is predictable.
The role of tropical-Pacific SST variability in driving multi-year NA rainfall variation is further explored with the Pacific Ocean-Global Atmosphere (POGA) pacemaker experiments from the GFDL-CM2.1 coupled model. The fully-coupled model is run globally but the SST anomalies in the equatorial eastern Pacific are restored to follow observed SST for the period 1900-2014. The ongoing work is untangling the potential contributions of tropical-Pacific SST for generating multi-year variations of NA rainfall, along with analysis of the pre-industrial control and the historical runs to provide more insight into the role of external forcing and internal noise for generating such multi-year rainfall variability.