Peter B. Gibson

Extreme heat in New Zealand is projected to intensify faster than mean warming, but the mechanisms remain unclear. Using dynamically downscaled CMIP6 projections, we investigate whether land–surface moisture constraints contributes to future heat‐extreme intensification in a maritime climate. Across six downscaled models, evaporative fraction (EF), a measure of surface energy partitioning, exhibits a robust negative relationship with the hottest day of the year (TXx), with the strongest coupling occurring approximately 2 months prior to TXx days. This highlights the importance of antecedent land‐surface conditions and land‐surface memory in preconditioning extreme heat. Future projections show widespread reductions in EF across New Zealand. These EF reductions are associated with additional TXx amplification of approximately 1°C–2°C across much of the North Island by the late 21st century. In contrast, projected circulation changes show limited spatial consistency across models. Our results demonstrate that land–atmosphere feedbacks can substantially amplify extreme heat even in maritime climates.

Regional climate modelling plays a central role in understanding and projecting climate change across Australasia, a region characterised by complex topography, diverse climate regimes, and high exposure to climate extremes. Over the past fifteen years, CORDEX-Australasia has coordinated regional climate downscaling under an international framework while operating alongside complementary national and state-based initiatives. Here we synthesise the evolution, achievements, scientific advances, applications, and future priorities of regional climate modelling in the region. Successive CORDEX-CMIP5 and CORDEX-CMIP6 ensembles have produced the largest coordinated high-resolution projection datasets for Australasia. Regional climate models demonstrate clear added value over their driving global climate models, particularly in representing temperature extremes, heavy precipitation, coastal processes, and complex terrain. CMIP6-based projections show robust agreement on substantial 21st-century warming, with uncertainty dominated by emissions scenarios, while precipitation responses remain more regionally heterogeneous. Advances in convection-permitting modelling, benchmarking frameworks, and bias-correction methods have further improved process representation and evaluation. Regional projections now underpin applications in water resources, agriculture, disaster risk reduction, energy, health, and national climate services. Progress has been supported by shared infrastructure through the National Computational Infrastructure and strengthened coordination via the National Partnership for Climate Projections, highlighting strong synergies between State and Federal programs contributing to CORDEX. Key challenges remain including computational demands, persistent model biases, precipitation uncertainty, ensemble design, and translating complex multi-model information into decision-ready guidance. Strategic priorities include kilometre-scale (convection-permitting) modelling, enhanced Earth system coupling, coordinated super-ensembles, responsible integration of artificial intelligence approaches, strengthened observational constraints, and expanded support for vulnerable remote Australian islands and Pacific Island nations. Sustained investment, collaboration, and innovation will be essential to deliver robust, high-resolution climate information for adaptation and resilience across Australasia.

Skillful seasonal forecasts of impactful climate events are in high demand yet remain a major challenge. Skill on seasonal timescales will arise from prediction of slowly varying predictable modes and their influence on the weather. The stratosphere provides one such mode, if it can be predicted, and if it exerts a sufficient influence on the troposphere below. Here we introduce a new seasonal hindcast data set with the Community Earth System Model version 2 (CESM2) with a raised atmospheric model lid and enhanced vertical resolution and compare with a companion hindcast set with CESM2's default vertical resolution. The enhanced vertical resolution enables considerably higher prediction skill of the Quasi‐biennial Oscillation (QBO). We further use this data set to probe for associated skill in features that are thought to be influenced by the QBO and use it to inform on the uncertainties that could be present in observed QBO connections given the short observational record. This includes connections to the Northern Hemisphere polar vortex, the North Atlantic Oscillation, the Madden Julian Oscillation, the sub‐tropical and mid‐latitude westerlies in the Pacific sector, tropical precipitation, and the tropical easterly jet. The enhanced QBO skill has little impact on these phenomena. However, the hindcasts do indicate an improved representation of connections between the QBO and the stratospheric polar vortex and tropospheric zonal winds in the Pacific sector with higher resolution but there are indications of a low signal‐to‐noise problem in QBO teleconnections that, if addressed, could improve prediction skill in some of these features.

Reliable projections of tropical cyclones (TCs) and associated impacts remain hampered by both climate model resolution and simulation length. To address this, here we present updated projections of TCs for the southwest Pacific from a high‐resolution downscaled ensemble of CMIP6 models. The downscaling implements a variable‐resolution atmospheric model enhancing resolution over the southwest Pacific and New Zealand (∼12–30 km). We assess future changes in TC frequency, changes in large‐scale environmental conditions, and associated extreme precipitation and winds across tropical and ex‐tropical storm phases. Changes in TC track pathways are also investigated through cluster analysis. Across the downscaled simulations, robust changes in TC frequency were not found, including for a high‐emissions scenario at end‐of‐century. Projections of the background environmental conditions are shown to be a significant source of uncertainty, owing to diverging projections of relative SST and tropical convection across the region in the host GCMs. However, very strong TCs (category 4 and above) show greater consensus for an increase in frequency, with 16 of 18 simulations across models and scenarios projecting an increase. Cluster analysis of TC tracks indicates a slight decrease in tracks that often impact northern parts of Australia. Extreme precipitation associated with TCs under a high‐emissions scenario is projected to increase by ∼30%–35% averaged across models, both for storms in the tropics and ex‐TCs impacting New Zealand. This increase exceeds Clausius‐Clapeyron scaling in five of six simulations. These projected increases in associated extreme precipitation pose significant societal risks despite the remaining uncertainty in TC frequency changes.

Orographic precipitation is a critical freshwater source and major flooding hazard, but its distribution and behavior over complex terrain are often uncertain due to sparse observations. We examine precipitation and its drivers in one of the wettest regions in the world, the Southern Alps of New Zealand (NZ), using the first multi‐decadal simulation by a convection‐permitting regional climate model across all mainland NZ at 2.2 km grid‐scale. Model skill is primarily assessed against direct measurements by more than 170 rain gauges to avoid uncertainty commonly introduced by gridded observations in remote regions. Peak intensity and duration of sub‐daily rainfall over mountains appear markedly improved in the 2.2 km model relative to the 12 km driving model. The orientation of water vapor flux relative to the mountain barrier strongly affects both climatological and daily extreme precipitation. Transects illustrate the influence of steep local topography on strong landfalling atmospheric rivers to produce high vertical velocities and extremely high accumulations of rainfall over windward upper mountain flanks, which do not appear unreasonable against available gauge observations. These transects also reveal the finer spatial structure of mountain waves in the 2.2 km model, which may contribute to its more realistic windward enhancement of orographic precipitation, but with excessive leeward precipitation and an annual mean dry bias over mountains. Despite the computational burden, these results support further targeted dynamical modeling at kilometer scales to improve physical understanding of precipitation in the current climate and its potential future change in NZ and other mountainous regions of the world.

Atmospheric rivers (ARs) play a critical role in moisture transport across the Southern Hemisphere mid‐latitudes and often produce extreme rainfall events across New Zealand. Here we examine ARs in a new set of high‐resolution (12–30 km) dynamically downscaled simulations from select CMIP6 models. We begin with a historical evaluation of AR properties from this model ensemble followed by future projections under a high emissions scenario. We demonstrate that by end‐of‐century, the maximum integrated vapor transport associated with ARs will robustly increase by as much as in some regions of the South Pacific, and ARs contribute a larger proportion (up to more) of the annual precipitation climatology in certain regions of New Zealand. The spatial structure and seasonality of these changes indicate the role of a poleward shift and intensification of the westerly jet. This is further quantified through a decomposition into dynamic and thermodynamic components. Although the thermodynamic change dominates the increase in AR frequency, the additional positive dynamical change is notable over the mid‐latitudes and southern New Zealand. Separating events using an AR categorization scale shows that higher category (longer duration, more intense) events more than double in frequency, underscoring the increasing role that ARs will play in extreme weather events in the future for this region.

Reliable projections of tropical cyclones (TCs) and associated impacts remain hampered by both climate model resolution and simulation length. To address this, here we present updated projections of TCs for the southwest Pacific from a high-resolution downscaled ensemble of CMIP6 models. The downscaling implements a variable-resolution atmospheric model enhancing resolution over the southwest Pacific and New Zealand (~12-30km). We assess future changes in TC frequency, changes in large-scale environmental conditions, and associated extreme precipitation and winds across tropical and ex-tropical storm phases. Changes in TC track pathways are also investigated through cluster analysis. Across the downscaled simulations, robust changes in TC frequency were not found, including for a high-emissions scenario at end-of-century. Projections of the background environmental conditions are shown to be a significant source of uncertainty, owing to diverging projections of relative SST and tropical convection across the region in the host GCMs. However, very strong TCs (category 4 and above) show greater consensus for an increase in frequency, with 16 of 18 simulations across models and scenarios projecting an increase. Cluster analysis of TC tracks indicates a slight decrease in tracks that often impact northern parts of Australia. Extreme precipitation associated with TCs under a high-emissions scenario is projected to increase by ~30-35% averaged across models, both for storms in the tropics and ex-TCs impacting New Zealand. This increase exceeds Clausius-Clapeyron scaling in 5 of 6 simulations. These projected increases in associated extreme precipitation pose significant societal risks despite the remaining uncertainty in TC frequency changes.

Anticipating climate impacts and risks in present or future climates requires predicting the statistics of high‐impact weather events at fine‐scales. Direct numerical simulations of fine‐scale weather are computationally too expensive for many applications. While deterministic‐based (deep‐learning or statistical) downscaling of low‐resolution climate simulations are several orders of magnitude faster than direct numerical simulations, it suffers from several limitations. These limitations include the tendency to regress to the mean, which produces excessively smooth predictions and underestimates the magnitude of extreme events. They also fail to preserve statistical measures that are key for climate research. We use a conditional GAN (cGAN) architecture to downscale daily precipitation as a Regional Climate Model (RCM) emulator. The cGAN generates plausible residuals on top of the predictable expectation state produced by a deterministic deep learning algorithm. The skill of cGANs is highly sensitive to a hyperparameter known as the weight of the adversarial loss (), where the value of required for accurate results varies with season and performance metric, casting doubt on the reliability of cGANs as usually implemented. However, by applying a simple intensity constraint to the loss function, it is possible to obtain reliable performance results across spanning two orders of magnitude. CGANs are considerably more skillful in capturing climatological statistics, including the distribution and spatial characteristics of extreme events. With this modification, we expect cGANs to be readily transferable to other applications and time periods, making them a useful weather generator for representing extreme event statistics in present and future climates.

While deep‐learning downscaling algorithms can generate fine‐scale climate projections cost‐effectively, it is unclear how effectively they extrapolate to unobserved climates. We assess the extrapolation capabilities of a deterministic Convolutional Neural Network baseline and a Generative Adversarial Network (GAN) built with this baseline, trained to predict daily precipitation simulated by a Regional Climate Model (RCM) over New Zealand. Both approaches emulate future changes in annual mean precipitation well, when trained on historical data, though training on a future climate improves performance. For extreme precipitation (99.5th percentile), RCM simulations predict a robust end‐of‐century increase with future warming (∼5.8%/C on average from five simulations). When trained on a future climate, GANs capture 97% of the warming‐driven increase in extreme precipitation compared to 65% in a deterministic baseline. Even GANs trained historically capture 77% of this increase. Overall, GANs offer better generalization for downscaling extremes, which is important in applications relying on historical data.