Ji et al. (2026) Three generations of NARCliM: future projections of mean and extreme climate over the CORDEX Australasia domain

Identification

• Journal: npj Climate and Atmospheric Science
• Year: 2026
• Date: 2026-01-14
• Authors: Fei Ji, Moutassem El Rafei, Giovanni Di Virgilio, Jason P. Evans, Jatin Kala, Stephen White, Julia Andrys, Dipayan Choudhury, Eugene Tam, Yue Li, Rishav Goyal, Carlos Rocha, Matthew Riley
• DOI: 10.1038/s41612-025-01280-w

Research Groups

• Climate & Atmospheric Science, NSW Department of Climate Change, Energy, the Environment and Water, Sydney, Australia
• Climate Change Research Centre, University of New South Wales, Sydney, Australia
• Australian Research Council Centre of Excellence for 21st Century Weather, University of New South Wales, Sydney, Australia
• Environmental and Conservation Sciences, and Centre for Climate Impacted Terrestrial Ecosystems, Harry Butler Institute, Murdoch University, Murdoch, WA 6150, Australia

Short Summary

This study evaluates future changes in mean climate and 10 extremes using three generations of the NARCliM project, which dynamically downscale CMIP3, CMIP5, and CMIP6 models. Projections consistently show statistically significant increases in maximum and minimum temperatures across all NARCliM generations, while precipitation projections exhibit greater variability, emphasizing the robustness of temperature extremes compared to precipitation.

Objective

• To systematically compare future climate projections from three generations of NARCliM (N1.0, N1.5, N2.0).
• To examine similarities and differences in projected changes to mean climate and extremes.
• To assess how updates in driving Global Climate Models (GCMs), emission scenarios, and model resolution contribute to these differences.
• To evaluate the robustness of projections across generations for Australia.

Study Configuration

• Spatial Scale:
  • Outer domain: CORDEX Australasia region (50 km resolution for N1.0/N1.5, 20 km for N2.0).
  • Inner domain: Southeast Australia (10 km resolution for N1.0/N1.5, 4 km for N2.0).
• Temporal Scale:
  • Historical period for comparison: 1990–2009.
  • Future projection period for comparison: 2060–2079.
  • N1.0 simulations: Three 20-year time slices (1990–2009, 2020–2039, 2060–2079).
  • N1.5 and N2.0 simulations: Continuous runs from 1950 to 2100.

Methodology and Data

• Models used:
  • Regional Climate Model (RCM): Weather Research and Forecasting (WRF) model (v3.3 for N1.0, v3.6 for N1.5, v4.1.2 for N2.0).
  • NARCliM1.0: Dynamically downscaled 4 CMIP3 GCMs (CCCMA3.1, ECHAM5, MIROC3.2, CSIRO-mk3.0) under SRES A2 emission scenario.
  • NARCliM1.5: Dynamically downscaled 3 CMIP5 GCMs (ACCESS1.0, ACCESS1.3, CanESM2) under RCP4.5 and RCP8.5 emission scenarios.
  • NARCliM2.0: Dynamically downscaled 5 CMIP6 GCMs (ACCESS-ESM1-5, EC-Earth3-Veg, MPI-ESM1-2-HR, NorESM2-MM, UKESM1-0-LL) under SSP1-2.6, SSP2-4.5, and SSP3-7.0 emission scenarios.
  • Extreme climate indices: 10 selected Expert Team on Sector-Specific Climate Indices (ET-SCI) (TXx, TNn, DTR, TXge35, WSDI, CDD, CWD, R10mm, Rx1Day, R99p).
  • Statistical significance tests: t-tests (for temperature), Mann–Whitney U test (for precipitation), and Tebaldi et al. (2011) approach for ensemble mean agreement.
• Data sources:
  • Global Climate Model (GCM) outputs from CMIP3, CMIP5, and CMIP6.
  • Reanalysis datasets for historical baseline: NCEP (for N1.0), ERA-Interim (for N1.5), and ERA5 (for N2.0).

Main Results

• Mean Temperature: All NARCliM generations consistently project statistically significant increases in maximum and minimum temperatures across Australia. Warming is generally greater in central and northern regions compared to the south, including Tasmania. The magnitude of warming is primarily driven by the selected GCMs and emission scenarios; for example, N1.5 projects the highest warming under RCP8.5, while N2.0's high-emission warming falls between N1.0 and N1.5. Seasonal variations show the smallest increases in winter, with greater warming in spring and summer.
• Temperature Extremes: Extreme heat indices (TXx, TNn, TXge35, WSDI) are projected to increase consistently and significantly across all NARCliM generations and scenarios, with larger increases under higher emissions. TXx (annual maximum daily maximum temperature) and TXge35 (days with maximum temperature ≥ 35 °C) intensify most in northern Australia. WSDI (warm spell duration index) shows a north-to-south decreasing gradient. TNn (annual minimum daily minimum temperature) increases are generally smaller than TXx, indicating a stronger intensification of extreme heat events relative to cold extremes. Diurnal Temperature Range (DTR) changes vary and are mostly statistically insignificant.
• Mean Precipitation: Precipitation projections exhibit greater variability and uncertainty across generations, emission scenarios, and regions, with most changes being statistically insignificant. N1.0 projects wetter conditions in central and northern Australia but drier conditions in the southwest. N1.5 shows increased rainfall in central-western areas but drier conditions elsewhere. N2.0 indicates slight increases in northern Australia but significant drying in the east and southwest. Seasonal variations in precipitation are inconsistent across models and scenarios.
• Precipitation Extremes: Projections for precipitation extremes vary, though some trends are consistent. CDD (consecutive dry days), Rx1Day (maximum 1-day precipitation), and R99p (annual sum of daily precipitation > 99th percentile) are generally projected to increase. CWD (consecutive wet days) and R10mm (days with precipitation ≥ 10 mm) are generally projected to decrease. N2.0 specifically suggests shorter wet periods and fewer heavy rain days, but more intense very extreme rainfall (R99p). These changes are mostly statistically insignificant.
• Driving Factors: Future climate projections across all NARCliM generations are primarily determined by the characteristics of their driving GCMs and the emission scenarios used, with RCMs largely preserving the large-scale trends from GCMs.

Contributions

• Provides the first comprehensive comparison of future climate projections (mean and extremes) across three generations of the NARCliM project, which dynamically downscale CMIP3, CMIP5, and CMIP6 models.
• Highlights the relative robustness and statistical significance of temperature extreme projections compared to the higher variability and uncertainty in precipitation projections across different model generations and emission scenarios.
• Demonstrates how advancements in GCMs, RCMs (e.g., WRF v4.1 with 4 km convection-permitting resolution in N2.0), and scenario design (SSP pathways) influence regional climate projections.
• Emphasizes the critical role of a broader and more diverse selection of driving GCMs to improve confidence and reduce uncertainties in regional climate projections.
• Underscores the necessity for global climate modeling initiatives (e.g., WCRP WGCM) to provide essential sub-daily, three-dimensional GCM output variables to facilitate robust dynamical downscaling efforts.

Funding

• New South Wales Department of Climate Change, Energy, the Environment and Water (NARCliM2.0 dynamical downscaling project contributing to CORDEX Australasia).
• NSW Climate Change Adaptation Strategy.
• NSW Climate Change Fund for NSW and Australian Regional Climate Modelling (NARCliM) Project.
• National Computational Infrastructure (NCI) (supported by the Australian Government).
• Australian Research Council Centre of Excellence for 21st Century Weather (CE230100012).
• Climate Systems Hub of the Australian Government's National Environmental Science Program.

Citation

@article{Ji2026Three,
  author = {Ji, Fei and Rafei, Moutassem El and Virgilio, Giovanni Di and Evans, Jason P. and Kala, Jatin and White, Stephen and Andrys, Julia and Choudhury, Dipayan and Tam, Eugene and Li, Yue and Goyal, Rishav and Rocha, Carlos and Riley, Matthew},
  title = {Three generations of NARCliM: future projections of mean and extreme climate over the CORDEX Australasia domain},
  journal = {npj Climate and Atmospheric Science},
  year = {2026},
  doi = {10.1038/s41612-025-01280-w},
  url = {https://doi.org/10.1038/s41612-025-01280-w}
}