Storylines for Future Projections of Precipitation Over New Zealand in CMIP6 Models

Large uncertainty exists in the sign of long-term changes in regional scale mean precipitation across the current generation of global climate models. To explore the physical drivers of this uncertainty for New Zealand, here we adopt a storyline approach applying cluster analysis to spatial patterns of future projected seasonal mean precipitation change across CMIP6 models (n = 43). For the winter precipitation change signal, the models split roughly into two main groups: both groups have a very robust wet signal across the west coast of the South Island but differ notably in terms of the sign of precipitation change across the north of the North Island. These far north winter precipitation differences appear related to how far the Hadley cell edge and regional eddy-driven jet shift across the models relative to their historical positions. In contrast, for summer, most models have a markedly weaker and spatially non-uniform response, where internal variability often plays a large role. However, a small group of models predict a robust wet signal across most of the country in summer. This "wet model" group is characterized by a regional La Nina-like increase in high pressure shifted further to the south-east of New Zealand, associated with more frequent north-easterly flow over the country and accompanied by significant warming of local sea surface temperatures. This regional circulation response appears related to changes in stationary Rossby wave paths as opposed to changes in La Nina occurrence frequency itself. At the regional scale, future changes to mean precipitation under climate change could carry large societal consequences. Unfortunately, large uncertainties still exist on regional scales which may hinder climate change adaptation efforts. Here we explore and characterize these uncertainties across the latest generation of global climate models for the New Zealand region. Across the models, winter precipitation changes are shown to be much more consistent compared to summer precipitation changes. In winter, changes in the jet stream and Hadley cell edge positions in the models are important for determining the regional spatial patterns of precipitation change. In summer, internal variability uncertainty plays a larger role, models that predict robust wet changes across the country are associated with more north-easterly flow conditions in the future period. Changes to Rossby wave pathways appear important for setting up this regional circulation response in summer. Storylines are used to characterize and explain the main precipitation change patterns across models Spatial patterns of precipitation change are more robust in winter, inter-model differences relate to Hadley cell and jet changes Spatial patterns of precipitation change are less robust in summer, internal variability and Rossby wave pathway changes are important