New Zealand atmospheric river (AR) lifecycles are analyzed to examine the synoptic conditions that produce extreme precipitation and regular flooding. An AR lifecycle tracking algorithm, novel to the region, is utilized to identify the genesis location of New Zealand ARs: the location where moisture fluxes enhance and become distinct synoptic features capable of producing impactful weather conditions. Genesis locations of ARs that later impact New Zealand cover a broad region extending from the Southern Indian Ocean (90°E) into the South Pacific (170°W) with the highest genesis frequency being in the Tasman Sea. The most impactful ARs, associated with heavy precipitation, tend to originate from distinct regions based on landfall location. Impactful North Island ARs tend to originate from subtropical regions to the northwest of New Zealand, while impactful South Island ARs are associated with genesis over southeast Australia. The synoptic conditions of impactful AR genesis are identified with North Island ARs typically associated with a cyclone in the central Tasman Sea along with a distant, persistent low pressure off the coast of West Antarctica. South Island AR genesis typically occurs in conjunction with moist conditions over Australia associated with a zonal synoptic-scale wavetrain. The Madden–Julian oscillation (MJO) is examined as a potential source of variability that modulates New Zealand AR lifecycles. It appears that the MJO modulates AR characteristics, especially during Phase 5, typically bringing more frequent, slow moving ARs with greater moisture fluxes to the North Island of New Zealand.

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 uses. While regression-based (deep-learning or statistical) downscaling of low-resolution climate simulations is 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. Additionally, they also fail to preserve statistical measures that are key for climate research. We use a conditional GAN (c-GAN) architecture to downscale daily precipitation as a Regional Climate Model (RCM) emulator. The c-GAN generates plausible residuals on top of the predictable expectation state produced by a regression-based DL algorithm. The skill of c-GANs is highly sensitive to a hyperparameter known as the weight of the adversarial loss (\(\lambda_{adv\ }\)), and the value of \(\lambda_{adv}\) required for accurate results varies with season and performance metric, casting doubt on the robustness of c-GANs as usually implemented. But, by applying a simple intensity constraint to the loss function, it is possible to obtain robust performance results across \(\lambda_{adv}\) spanning two orders of magnitude. C-GANs are considerably more skillful in capturing climatological statistics including the distribution and spatial characteristics of extreme events. We expect c-GANs with this modification to be readily transferable to other problems and time periods, making them a useful weather generator for representing extreme event statistics in present and future climates.