Zhou et al. (2026) Predicting the unprecedented: assessing contributions from large-scale modes of variability and climate change to Southeast Australia’s record spring rainfall in 2022

Identification

• Journal: Climate Dynamics
• Year: 2026
• Date: 2026-04-01
• Authors: Linjing Zhou, Eun-Pa Lim, Pandora Hope, Griffith Young, Acacia Pepler, I. SIMMONDS
• DOI: 10.1007/s00382-026-08122-2

Research Groups

• Bureau of Meteorology, Melbourne, VIC, Australia
• School of Geography, Earth and Atmospheric Sciences, The University of Melbourne, Parkville, VIC, Australia

Short Summary

This study quantifies the contributions of large-scale climate drivers and anthropogenic global warming to Southeast Australia's record spring 2022 rainfall. It reveals that while these factors explained a substantial portion, local atmospheric conditions and an increased frequency of intense weather systems played a critical role in amplifying the event's unprecedented extremity, with anthropogenic climate change contributing approximately 12% to the total rainfall.

Objective

• To quantify how well the state of La Niña, the negative Indian Ocean Dipole (IOD), and the positive Southern Annular Mode (SAM) explained the extent of the 2022 Southeast Australia (SEA) spring rainfall.
• To assess the predictability of the extreme rainfall in the dynamical seasonal forecast system.
• To identify other climate/weather features that contributed to the extremity of the spring 2022 rainfall and determine whether oceanic or atmospheric conditions were a more important source of predictability for SEA rainfall.
• To estimate the extent to which long-term climate change, driven by increasing greenhouse gas (GHG) concentrations, influenced the forecast climate anomalies of spring 2022.

Study Configuration

• Spatial Scale: Southeast Australia (30–45° S, 140–156° E), broader Australian continent, Southern Hemisphere extratropics, tropical Pacific and Indian Oceans.
• Temporal Scale: Austral spring (September–November) 2022 for the event; 1900–2022 for historical record; 1981–2018 for climatological baseline and model hindcasts; 1905 for counterfactual climate change experiments.

Methodology and Data

• Models used:
  • Multiple linear regression model
  • ACCESS-S2 (Australian Community Climate-Earth System Simulator – Seasonal version2)
  • ACCESS-MICAS (Modified Initial Condition Attribution System)
  • UK Met Office’s GloSea5 (component models)
  • Met Office Global Coupled model 2.0 (GC2)
• Data sources:
  • NOAA SST data (0.25 degree monthly, optimum interpolation of in-situ and satellite observations)
  • ERA5 (European Centre for Medium-Range Weather Forecasts Re-Analysis, monthly mean data)
  • Australian Gridded Climate Data (AGCD) version 2 for SEA rainfall
  • Bureau of Meteorology’s 4-D Var data assimilation analyses (for ACCESS-S2 real-time forecasts)
  • ERA-Interim reanalysis (for ACCESS-S2 hindcasts)
  • CMIP5 forcings (for GC2 climate model)
  • Cyclone tracking dataset (University of Melbourne tracking scheme)
  • Front tracking dataset

Main Results

• Southeast Australia experienced its wettest spring in the 126-year record in 2022, with rainfall almost twice the 1981–2018 average and 60% higher than the previous record.
• Large-scale climate drivers (moderate La Niña, negative IOD, strong positive SAM) and anthropogenic global warming explained a substantial fraction, but not the full magnitude, of the record rainfall.
• Anthropogenic climate change contributed approximately 12% to the total rainfall in spring 2022, linked to generally higher sea surface temperatures (north of Australia, Tasman Sea) and a stronger positive SAM in the present climate.
• Both the multiple linear regression model and the ACCESS-S2 ensemble mean forecasts underestimated the extreme magnitude of the rainfall, predicting it as the second wettest (statistical) or less extreme than the 2010 event (ACCESS-S2 ensemble mean).
• Anomalous atmospheric conditions played a dominant role in amplifying the event's extremity; experiments with random atmospheric initial conditions (RandALIC) showed an approximately 80% reduction in rainfall compared to control forecasts.
• The wettest ACCESS-S2 ensemble member, which simulated rainfall comparable to observations, also simulated a stronger positive SAM and polar vortex, along with a zonal wavenumber-1 pattern featuring intense high-pressure anomalies south of Australia.
• The spring-mean Thermal Wind Index (TWI) in 2022 was the strongest on record, correlating significantly with SEA rainfall, but this relationship largely depended on its co-variability with ENSO, IOD, and SAM. The residual TWI (after removing large-scale drivers) was still record-large, suggesting a role for local factors.
• The number of rain-bearing cyclones was record-high, and the number of wet fronts was the third highest on record in spring 2022, both contributing to the high rainfall and the record TWI.

Contributions

• Quantified the relative contributions of large-scale climate modes of variability and anthropogenic climate change to an unprecedented regional extreme rainfall event using both statistical and state-of-the-art dynamical forecast-based attribution systems.
• Demonstrated the utility of the ACCESS-S2 operational dynamical forecast system and its climate change attribution system (ACCESS-MICAS) for understanding and attributing extreme events on sub-seasonal to seasonal timescales.
• Highlighted that while large-scale drivers and climate change set a conducive environment, the extreme magnitude of the 2022 rainfall was significantly amplified by anomalous atmospheric conditions and specific characteristics of mid-latitude weather systems (e.g., record-strong thermal wind, increased frequency of wet cyclones/fronts), indicating intrinsically limited predictability for such events.
• Provided insights into the complex, potentially non-linear interactions between oceanic, atmospheric, and terrestrial processes, and increasing GHGs in shaping extreme rainfall events.

Funding

• National Environmental Science Program (NESP), funded by the Australian Government’s Department of Climate Change, Energy, the Environment and Water.
• Victorian Water and Climate Initiative (VicWaCI), funded by the Department of the Energy, Environment, and Climate Action.
• National Computational Infrastructure (NCI).

Citation

@article{Zhou2026Predicting,
  author = {Zhou, Linjing and Lim, Eun-Pa and Hope, Pandora and Young, Griffith and Pepler, Acacia and SIMMONDS, I.},
  title = {Predicting the unprecedented: assessing contributions from large-scale modes of variability and climate change to Southeast Australia’s record spring rainfall in 2022},
  journal = {Climate Dynamics},
  year = {2026},
  doi = {10.1007/s00382-026-08122-2},
  url = {https://doi.org/10.1007/s00382-026-08122-2}
}