Intensification of Short-Duration Extreme Precipitation in Greater Sydney

Short-duration extreme precipitation is a key driver of urban flooding and associated socio-economic impacts in a warming climate. Increasing urbanization further amplifies the vulnerability of cities to intense rainfall occurring over minutes to hours. These extremes frequently trigger flash floods and pose substantial risks to urban infrastructure and public safety. Despite growing recognition of its importance, regional-scale assessments of sub-hourly extreme precipitation remain limited. Emerging observational evidence indicates that short-duration precipitation events (≤1 hour) are intensifying at a faster rate than longer-duration events. In this study, we analyze short-duration extreme precipitation events at 5-, 10-, 20-, 30-, and 60-minute timescales using observations from 16 automated weather stations (AWS) across the rapidly urbanizing Greater Sydney region, New South Wales, Australia. Our results show a pronounced increasing trend in extreme precipitation at higher percentiles, particularly at the 5–10 minute timescales, compared to hourly extremes. At the hourly scale, we evaluate the performance of five convection-permitting regional climate model simulations (4 km ensemble) against AWS observations. The models reasonably capture the upper tail of the precipitation distribution but tend to slightly overestimate the frequency of extreme events. To assess future changes, we examine the intensity of 99^th percentile precipitation extremes across three periods—historical (1951–2014), near future (2015–2057), and far future (2058–2100)—under three Shared Socioeconomic Pathway scenarios (SSP126, SSP245, and SSP370). The projections indicate a consistent intensification of extreme precipitation, with a substantial upward shift in the top 1% of historical extremes, most pronounced under the high-emission SSP370 scenario. Interestingly, the simulations also project a reduction in the total number of wet hours relative to the historical baseline, suggesting a transition toward shorter-duration but more intense precipitation events. Although considerable inter-model spread and spatial variability exist, increases in 99^th percentile extremes are robust across most scenarios. Notably, under SSP126, a decline in extreme precipitation is projected in the far future, highlighting the potential benefits of strong emission mitigation. These findings underscore the need to explicitly incorporate short-duration precipitation extremes into urban planning and flood risk management under climate change.

Keywords: Automatic Weather Station, Climate change, Flash floods, NARCliM2.0, Regional climate models, Sub-hourly extreme precipitation