The rapid and unprecedentedwarming of the climate system has led to an increase in the frequency and severity of extreme weather events, presenting significant challenges to ecosystems, societies, and economies across the globe. In Aotearoa New Zealand (hereafter, ANZ), a country with an intricate topography and distinct ecology, the consequences of these severe extremes are evident. Hence it is critically important to understand how human-induced forcings are shaping and amplifying the frequencies and intensities of extremes within the country.

We use a large ensemble generated over the ANZ domain using the weather@home (w@h) regional climate model under two forcing scenarios. The first scenario represents pre-industrial settings, without any human-induced alterations, while the second scenario represents present-day conditions, with anthropogenic influences. Synoptic circulation patterns play a key role in shaping the weather of ANZ, thereby influencing the frequency and intensity of extreme weather events, ranging from heavy rainfall to droughts. In this thesis, we aim to statistically quantify how extremes such as temperature and precipitation have evolved with human-induced forcings over ANZ in combination with various large-scale circulation patterns.

Self-organising maps, a machine learning classification approach was employed to derive the 12 major synoptic circulation patterns around ANZ. These circulation types represent a range of pressure patterns and are the basis of all the analysis in this thesis.

In ANZ, due to anthropogenic forcings, there is an average increase of two to three times in the frequency of temperature extremes, based on the analysis of the w@h data. An approximate 1◦C rise in the mean temperature is also observed in the ANT scenario compared to the NAT scenario. Although certain synoptic states like low-pressure systems over the east/northeast of ANZ are infrequent, the frequency of temperature extremes is increased by a factor of seven in the ANT scenario than NAT during the winter season. Low-pressure centres to the northwest of New Zealand are frequently associated with high temperatures in both the ANT and NAT scenarios, likely linked to the advection of warm air of tropical origins for these patterns. The frequency of extreme temperatures in these synoptic patterns has also doubled between theNAT and ANT ensembles. In both scenarios, regions of the North Island like Northland and the east coast of the ANZ such as Canterbury regions experience higher temperatures with these circulation types. However, the west coast along the Southern Alps in the South Island experiences the largest increase in extreme temperatures with anthropogenic forcings.

With extreme temperatures rising in intensity and frequency across all regions of ANZ due to anthropogenic influences, changes in precipitation patterns are anticipated, following the Clausius-Clapeyron (C-C) relationship. Therefore, this analysis examines the effects of human-induced forcing on precipitation within various large-scale circulation types. We also quantify the thermodynamic and dynamic influences of different circulation types on precipitation over ANZ relative to C-C rates. In general, the increase in precipitation intensity due to human-induced warming is less than the anticipated C-C value at the lower percentiles, but aligns with the C-C rates at higher percentiles for extreme precipitation. This is particularly clear for low-pressure systems to the southwest of ANZ and associated with westerly winds onto the South Island. These nodes also exhibit higher precipitation change due to human-induced warming, particularly in the west coast regions. Whereas, northwesterly circulation type display a reduction in precipitation in the ANT scenarios especially over the North Island compared to the expected C-C value. In addition, the number of wet days decreases in the ANT scenario compared to the NAT scenario. The highest reductions are observed on the west coast of the South Island for westerly winds, while there are slight increases on the east coast. However, the frequency of days with extreme precipitation increases for most circulation patterns, except for Northland and northwesterly flows. This highlights the combined impact of dynamics and thermodynamics in defining changes in both the intensity and occurrence patterns of precipitation across ANZ.

Finally, we examine the changes in extreme temperatures, when coupled with dry, wet, and humid conditions, attributed to human-induced forcing over ANZ. Furthermore,we explore the impact of large-scale synoptic features on these alterations. High-pressure circulation types display a higher frequency of dry days and dry hot extremes, although their increase under the ANT scenario is less pronounced compared to the rise observed for wet hot extremes and humid hot extremes. Extreme temperatures tend to intensify on dry days, whereas they decrease on wet and humid days. While high relative humidity is less frequent and is typically associated with lower temperatures, its combination with extreme temperatures can exacerbate their severity. In the eastern coast regions of the South Island, hot humid temperatures intensify by over 4◦C due to human-induced warming in the ANT simulations during westerly flow patterns. This underscores the significant impact of relative humidity on extreme temperatures despite frequent dry hot days caused by the westerlies and the associated foehn effect in these regions.