Substantial increases in future precipitation extremes – insights from a large ensemble of downscaled CMIP6 models

Substantial increases in future precipitation extremes – insights from a large ensemble of downscaled CMIP6 models | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Substantial increases in future precipitation extremes – insights from a large ensemble of downscaled CMIP6 models Rohan Eccles, Jozef Syktus, Ralph Trancoso, Sarah Chapman, Conrad Wasko, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6291947/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Extreme precipitation events are widely held to become more intense and frequent as a result of climate change, which will have major impacts for future flooding with implications for the environment, infrastructure, agriculture, and human life. We investigated projected changes to daily mean, moderately extreme (99th and 99.7th percentile), and rare extreme (Annual Exceedance Probability (AEP) 1 in 10, 50, and 100) precipitation events across Australia and its greater capital cities, where approximately two thirds of the Australian population reside. We used dynamically downscaled CMIP6 precipitation simulations from 4 modelling groups in Australia. This large ensemble consists of 19 different host models downscaled using 3 distinct regional climate models in 5 different configurations, making an ensemble of 39 different downscaled simulations. The changes in mean and extreme precipitation events were quantified at each grid cell from each of the models according to the rate of change per degree of global warming. The largest increases to precipitation extremes were seen over northern Australia, with the 1 in 100 AEP event in Darwin projected to increase by 11.9% K − 1 and 12.2% K − 1 for the downscaled and host ensemble averages, respectively. Other capital cities had lower increases but still substantial (7.6% K − 1 for Brisbane, 7.3% K − 1 for Sydney, 3.4% K − 1 for Melbourne, and 4.4% K − 1 for Perth). Large spatial differences were noted among the downscaled ensembles, with models from different modelling groups showing varying spatial patterns and magnitudes of change. These results highlight the influence of the downscaling approach in determining changes to precipitation extremes and show the need to consider large ensembles to ensure uncertainties in host models and downscaling methods can be accounted for. The findings can inform decision making around flood management, urban planning, urban water supply and agriculture around Australia, in addition to revealing globally relevant scientific insights. Earth and environmental sciences/Climate sciences/Climate change/Projection and prediction Earth and environmental sciences/Climate sciences/Hydrology Earth and environmental sciences/Natural hazards Downscaled climate projections Extreme precipitation uncertainty Generalised Extreme Value distribution Global Warming Levels Precipitation extremes Regional climate modelling Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 1 Introduction The intensity and frequency of extreme precipitation events is widely expected to increase as a result of climate change 1 – 4 , which will have significant implications for future flooding 5 , 6 . Globally, observations show an increase in annual maximum daily precipitation between 5.9% and 7.7% per degree of warming 7 . Even larger increases have been reported for the 99th (11% K − 1 ) and 99.97th (13% K − 1 ) percentile precipitation 8 . Observed changes are a result of thermodynamic and dynamic processes, both of which are influenced by a warming atmosphere 9 . Thermodynamic processes such as the Clausius-Clapeyron (CC) relationship, which increases the water-holding capacity of the atmosphere by 6–7% per degree of warming 9 lead to relatively spatially homogeneous changes in precipitation intensity on the order of 4–8% per degree of warming 10 . Daily precipitation extremes have been shown to be intensifying at a rate approximately equal to the CC relationship 11 . Dynamic processes substantially modify the regional response to climate warming, through changes to the frequency and intensity of synoptic and subsynoptic phenomena, including tropical and extratropical cyclones 10 . Increased warming also increases atmospheric stability, which weakens circulation, potentially reducing the intensity of precipitation extremes. On the other hand, latent heat release results in atmospheric instability that can strengthen storms and precipitation extremes, with the most extreme events seeing the largest increases 12 , 13 . The overall impact of climate change on precipitation intensity is therefore highly complex and regionally dependent. Climate models are currently the best physically based approaches available to understand future precipitation processes, characteristics, and impacts. Global Climate Models (GCMs) from the latest Coupled Model Intercomparison Project Phase 6 (CMIP6) have shown increases to precipitation extremes in-line with the CC relationship, with smaller increases noted for more moderate events compared to rarer events 13 – 16 . Abdelmoaty & Papalexiou 17 showed historic 1 in 100 Annual Exceedance Probability (AEP) precipitation events would occur approximately every 50 and 70 years for the Northern and Southern Hemispheres, respectively by mid-century (2035–2067). Using 25 GCMs, Gründemann et al. 15 found the intensity of 1 in 100 AEP precipitation events increased by between 13.5% and 38.3%, depending on the emissions scenario, with smaller increases for more frequent events. An ensemble of CMIP6 GCMs projected increases to 1 in 10 and 1 in 50 AEP daily and 5-daily precipitation at rates approximately equal to the CC relationship (~ 7% K − 1 ) 16 . GCMs are, however, applied at relatively coarse resolutions (~ 150 km) and have difficulty adequately representing precipitation patterns over complex terrain and resolving extreme precipitation processes 18 . These coarse resolution models are not always suited to providing reliable regionally specific information required to support adaptation and decision-making at regional scales. To overcome these issues, Regional Climate Models (RCMs) dynamically downscale GCMs to a finer resolution (typically 10 to 50 km) in order to better represent small-scale features and processes. RCMs have been shown to have improved skill in representing patterns of local precipitation and the impacts of topography, coasts, and land use changes compared to GCMs 19 – 22 and may also be more skilful in simulating extremes 20 , 23 , 24 . In Australia, most studies to date using RCMs have focused on percentiles or used extreme indices defined by the Expert Team on Climate Change Detection and Indices 25 , such as the RX1day or the 99th percentile 26 – 28 . While these moderate extreme events are important, most flood impacts are associated with rarer extremes, which are less well understood. Herold et al. 29 projected 1 in 20 AEP daily precipitation would occur 1–2 times more frequently in capital cities across southeast Australia using projections from 4 downscaled CMIP3 GCMs, though there was low model agreement for these changes. Mantegna et al. 30 showed daily intensities could increase by over 15% K − 1 for the most extreme events (1 in 100 AEP) over Tasmania using 5 downscaled GCMs, with smaller increases for more moderate events. Similarly, Wasko, Guo et al. 31 found larger increases for more extreme daily events (up to the 1 in 50 AEP) compared to less extreme events across Australia using 4 statistically downscaled GCMs, with larger increases for the north and smaller increases for the southwest. There are significant uncertainties associated with projections of precipitation from climate models 32 , which necessitates large ensembles to better account for this uncertainty, especially when evaluating extremes 33 . However, studies on rare extremes to date have largely adopted limited ensemble sizes and used earlier climate projections, which may underestimate the uncertainty. There is a need to determine how climate change influences precipitation extremes using a large ensemble of the latest downscaled CMIP6 projections, particularly for the rarest events. The focus of this paper is therefore to examine the impacts of climate change on precipitation extremes using a large ensemble of downscaled CMIP6 projections. For this purpose, we use downscaled ensembles from 4 modelling groups under a high emission Shared Socioeconomic Pathway (SSP370). This large ensemble forms the core of Australia’s new national projections, contributing to the CORDEX-CMIP6 Australasia domain 20 , 34 – 36 , and underpins Australia’s climate services and adaptation planning. This large ensemble consists of 19 different host GCMs downscaled using 3 distinct regional climate models in 5 configurations, for a total of 39 different downscaled simulations (Table 1 ; more details in methods section). This represents the largest ensemble of high-resolution climate projections used to assess changes to precipitation in Australia, allowing for a more robust assessment of the uncertainty and the spatial variability of likely changes. The inclusion of multiple different downscaled ensembles also allows for an assessment of the uncertainty introduced from the downscaling approach, which has not previously been quantified for rare precipitation extremes. We focus on greater capital cities as this is where two thirds of the Australian population lives and thus where the impacts are likely to be most pronounced. The aims of this paper are threefold: to assess the performance of the CMIP6 GCMs and different RCM ensembles in simulating mean and moderately extreme precipitation (i.e., 99th and 99.7th percentiles). to investigate changes and uncertainty in mean, moderately extreme (99th and 99.7th percentiles), and rare extreme precipitation (1 in 10, 50, and 100 AEP) per degree of global warming based on a large ensemble of regional models. to investigate how extreme precipitation varies spatially across the continent and identify regions with more pronounced extreme precipitation increases. 2 Results 2.1 Evaluation of simulated daily mean and moderate extreme precipitation We assess the performance of the host GCMs and the 4 downscaled ensembles in simulating daily mean and moderately extreme precipitation across Australia by comparing to observation data. The ensemble of CMIP6 GCMs performed reasonably well in representing mean and moderate extreme precipitation across Australia, with lower average errors over Australia than some of the downscaled ensembles (Fig. 1 ). The performance of the CMIP6 ensemble was, however, lacking over coastal and mountainous regions, with a consistent dry bias, particularly for the moderate extremes (p99 and p99.7). This is evident when evaluating biases for the greater capital city regions, which are all situated near to the coast and/or exhibit complex terrain. In these regions, where the majority of the population resides, the CMIP6 GCMs were shown to perform significantly worse in representing moderate extremes compared to the downscaled ensembles, though performance was comparable when evaluating mean precipitation (Fig. 2 ). Of the downscaled ensembles, NARCliM2.0 and QldFCP-2 had notable dry biases over northern and western Australia, with smaller biases over southeast Australia (Fig. 1 ). By contrast, CCAM-ACS was noted to have an opposing wet bias that was greatest in the interior, including for southeast Australia. BARPA-ACS was the best performing ensemble over Australia as a whole for both mean and moderately extreme precipitation, with minimal MAPE and RMSE values. Within the greater capital city regions, all downscaled ensembles reported lower biases for p99.7 compared to mean precipitation and p99, except for CCAM-ACS, which had lower biases for mean precipitation (Table S1 to S3). For mean precipitation, CCAM-ACS had the lowest bias of any ensemble in the capital cities (Table S1 ). QldFCP-2 and NARCliM2.0 both showed a clear dry bias over most capital cities for mean precipitation, whereas biases for CCAM-ACS and BARPA-ACS varied according to the city. For p99 and p99.7, the performance of all downscaled ensembles varied according to location (Fig. 1 ). For p99.7, CCAM-ACS tended to overestimate precipitation intensity in capital cities, whereas NARCliM2.0 tended to underestimate intensity, though biases for NARCliM2.0 were relatively small. In contrast to the mean precipitation, QldFCP-2 also tended to have low biases for p99.7 except for in Perth and Hobart where the models showed a clear dry bias. BARPA-ACS also had good agreement except for in Brisbane and Adelaide where a wet bias was noted (Table S3). These biases are, however, in almost all cases an improvement when compared to the CMIP6 host models for the capital cities, suggesting that the downscaled models from these ensembles may be better suited to represent precipitation extremes in these regions compared to the host models. 2.2 Projected changes to mean, moderate and rare extreme precipitation 2.2.1 Changes across Australia Considerable spatial variability was evident in the projections of future mean and moderately extreme precipitation over Australia (Fig. 3 ). Spatial patterns of change varied according to the ensemble applied, with a consensus towards wetting or drying evident in a few regions only. A clear drying pattern for mean precipitation was evident in southwest Western Australia and southern Victoria and Tasmania to a lesser extent. This, however, did not necessarily lead to clear decreases for moderately extreme precipitation over these regions. In other regions, a drying or wetting signal was evident according to the signal-to-noise ratio for some ensembles but not for others. For example, the QldFCP-2 ensemble shows widespread agreement for decreased mean and p99 along coastal regions of north-eastern Australia, while the BARPA-ACS ensemble shows widespread agreement for increases over northern Australia, neither of which are reflected in the outputs for the other modelling groups. Similarly, over southeast Australia all model ensembles except for NARCliM2.0 show widespread agreement for increased p99.7. These results indicate that relying on a single ensemble of downscaled projections may give a false sense of the certainty of precipitation changes. All model ensembles show precipitation to vary according to the metric assessed, with the smallest (greatest) increases (decreases) seen for mean precipitation, with greater increases noted for more intense events compared to less intense events. Ensemble mean precipitation was projected to change by between − 2.5% K − 1 and 2.6% K − 1 , while p99.7 changed by between − 0.1% K − 1 and 4% K − 1 , depending on the ensemble used (Fig. 3 ). These changes are greater when evaluating the response of rare extremes to climate change (Fig. 4 ), where ensemble median changes ranged from 2.6% K − 1 to 6.5% K − 1 for the 1 in 10 AEP event and from 4.5% K − 1 to 10.1% K − 1 for the 1 in 100 AEP event. Both NARCliM2.0 and QldFCP-2 projected smaller increases to rare precipitation extremes over Australia than the CMIP6 host models (Fig. 4 ). By contrast, CCAM-ACS and BARPA-ACS projected increases greater than the CMIP6 host models. The increases from most ensembles tended to be greatest over northern Australia, especially from the BARPA-ACS ensemble. Despite this inter-ensemble agreement, there were few areas in any ensemble which showed agreement for the 1 in 100 AEP according to the signal-to-noise ratio, due to the variability of future extremes. Corresponding changes to the spatial patterns of the GEV fitted parameter values are presented in the Supplementary Materials (Figure S1 ). Over Australia, the intra-model variability tended to be greater within the NARCliM2.0 ensemble than the other modelling groups, despite sampling the least GCMs (Fig. 3 ). This is especially evident for p99.7 and appears to be due to the very wet model (EC-Earth3-veg) and very dry (ACCESS-ESM) becoming more wet and dry following downscaling (Fig. 5 ). By contrast, the other downscaling methods tended to bring these outlier models in towards the mean, with some exceptions, including EC-Earth3 downscaled with BARPA-ACS which showed the largest increases of any model assessed for both the 99.7th percentile and the 1 in 100 AEP event (Fig. 5 ). The QldFCP-2 ensemble tended to have the least variability in the projections, followed by the CCAM-ACS. As both ensembles share the same model for downscaling, this could relate to the CCAM model effectively dampening the magnitude of changes from outlier models. 2.2.2 Changes within Capital Cities There was widespread agreement from all ensembles for a decrease to mean precipitation (-11.2% K − 1 to -8.6% K − 1 ), p99 (-5.6% K − 1 to -4.2% K − 1 ), and p99.7 (-3.6% K − 1 to -1.5% K − 1 ) in Perth (Fig. 6 ) with smaller decreases for more intense events compared to the mean. In other capital cities there was less consensus in the magnitude or sign of change, though generally decreases to mean precipitation were projected from most ensembles in Adelaide and Melbourne, with only CCAM-ACS tending towards a slight increase. Most capital cities tended towards an increase to the more intense events (p99.7) compared to the mean, with the largest increases typically noted for Darwin (-2.3% K − 1 to 8.3% K − 1 ). In comparison to the projections for the means and moderate extremes, rare extremes tended towards an increase for almost all ensembles in all greater capital cities. Darwin noted the largest ensemble median increases in the 1 in 100 AEP event (8.3% K − 1 to 18.2% K − 1 ), with the greatest increases reported from the BARPA ensemble (Fig. 7 ). Here, the BARPA ensemble appears to show greater projected changes to extremes compared to the other CORDEX ensembles (Fig. 7 ). Similarly, NARCliM2.0 appears to show lower projected changes in Melbourne, while QldFCP-2 shows smaller changes in Brisbane when compared to the other CORDEX ensembles. In general, there were considerable differences in the projections across all capital cities, particularly those in northern and eastern Australia (Brisbane, Darwin, Sydney, ACT, and Melbourne). This variability is evident when comparing the distribution of the combined downscaled CORDEX ensemble to the individual downscaled ensembles (Fig. 6 and Fig. 7 ), highlighting the uncertainty associated with projections of precipitation and the need to adopt large ensembles using different downscaling methodologies to quantify the uncertainty. Generally, the combined downscaled ensemble appears to have a similar spread of projected changes to the host model ensemble, except in those regions discussed above (Fig. 7 and Fig. 8 ). The median projection of change from the host model ensemble closely matched that from the weighted downscaled ensemble (each modelling group weighted evenly; Fig. 8 ). This was, however, not true for Adelaide where the host models showed 1 in 100 AEP was projected to increase by 9% K − 1 in Adelaide compared to 4% K − 1 from the weighted downscaled ensemble or Brisbane where the host models showed a 3.7% K − 1 compared to 7.6% K − 1 from the weighted downscaled ensemble. Both groups projected the largest increases to the 1 in 100 AEP for Darwin (12.2% K − 1 and 11.9% K − 1 for the host and downscaled ensembles, respectively), which is considerably larger than what would be expected from the CC relationship (Fig. 8 ). 3 Discussion Our study shows downscaling consistently reduces biases of moderately extreme precipitation (p99 and p99.7) over capital city regions in Australia when compared to the host CMIP6 GCMs (Fig. 1 and Table S2 to S3). In contrast, the host models performed comparably well for mean precipitation within the capital city regions and for all mean and moderately extreme precipitation when assessed across Australia as a whole. These differences relate to the coarse model resolution of the GCMs, which are not able to represent precipitation patterns and extremes over regions with complex terrain, land use changes, or coastal gradients when compared to finer resolved RCMs 18 – 20 . It is, however, within these complex regions where the majority of the Australian population resides and where the performance differences between host and downscaled models are greatest. RCMs therefore appear to be better suited to provide information on climate hazards within these populated areas, particularly for extremes. There were considerable differences in the spatial extent of biases from the different downscaled modelling ensembles. Both NARCliM2.0 and QldFCP-2 had notable dry biases over Australia and the capital city regions, while CCAM-ACS and BARPA-ACS had notable wet biases, though to a lesser extent. The dry biases from the QldFCP-2 ensemble are particularly evident over northern Australia and may relate to a misrepresentation of the number and intensity of low-pressure systems and cyclones, which are major contributors of precipitation, especially extremes in these regions 37 . NARCliM2.0 shows similar dry biases over northern Australia and within the greater capital city regions (Fig. 2 ) for mean and p99, but much less bias for p99.7. Di Virgilio et al. 34 posited that the dry bias over northern Australia could relate to issues in capturing the Australian monsoon. They also showed that there was less bias in mean precipitation over southeast Australia from NARCliM2.0 compared to previous downscaled models from CMIP5 (NARCliM1.5) and CMIP3 (NARCliM1.0). By contrast, the wet bias present in the CCAM-ACS ensemble has been shown to relate to an overestimation of extreme precipitation, while low precipitation events were underestimated from this ensemble 36 . These biases are believed to relate to the parameterisation of CCAM, with improved schemes currently being tested to resolve these issues. Similarly, the wet bias from the BARPA ensemble has been shown to relate to a general overestimation of precipitation extremes, and inclusion of too many low intensity wet days. It is interesting to note the different sign of the biases between the CCAM-ACS and the QldFCP-2 as both ensembles make use of the CCAM model for downscaling. These differences likely relate to the downscaling approach adopted, particularly in regard to sea surface temperatures, which have a significant influence on precipitation. Here, QldFCP-2 elected to bias correct sea-surface temperatures from the host models, whereas CCAM-ACS adopted a nudging approach for the atmosphere and sea-surface temperatures, leading to diverging simulations of precipitation from the same host model. Extreme precipitation was projected to increase across Australia and all capital cities, with greater increases seen for rarer extremes compared to more moderate extremes (Fig. 9 ), in line with previous studies 13 – 16 . The magnitude of the projected changes were, however, dependent on the combination of the model ensemble and region considered. We found that changes for the 99.7th percentile precipitation ranged from between − 0.1% K − 1 to 4.0% K − 1 (Fig. 3 ), while changes to the 1 in 100 AEP ranged from between 4.5% K − 1 to 10.1% K − 1 across Australia (Fig. 4 ), which are the result of increases to all GEV parameters (Figure S1 and Figure S2). These projected changes are broadly in line with the findings of Wasko, Westra, et al. 38 , who suggested a scaling rate of 8% K − 1 for Australia from a meta-analysis of available studies of observations and projections. To date, available studies on projections have made it difficult to ascertain if there are any geographic differences in these scaling rates 38 , as the majority of the literature to date has focused on the populated southeast Australia 29 , 30 . Using the largest ensemble of projections to date, we show that there does appear to be large-scale geographic differences in projected precipitation extremes, with greater increases generally projected for northern Australia compared to southern Australia. Similar findings have been noted when assessing the observational records 2 , 39 and from a recent study based on 4 downscaled CMIP5 GCMs 38 . While this north-south difference is generally evident across the ensembles, it is important to note that there are widespread regional variations from the different ensembles for all precipitation intensities, except for southwest Australia where a strong signal of decreasing mean precipitation is shown for all ensembles (Fig. 3 ). Decreased mean precipitation, however, did not necessarily translate to a decrease in moderate or rare extremes, with generally consistent increases to rare extremes still projected for Perth despite the declines to mean precipitation (Fig. 9 ). In other regions, a drying or wetting signal is evident according to the signal-to-noise ratio for some ensembles but not for others. For example, the QldFCP-2 ensemble reported significant decreases to mean and 99th percentile precipitation for coastal regions of north-eastern Australia, whereas BARPA-ACS noted significant increases over northern Australia, neither of which are reflected in other modelling group outputs. Relying on a single ensemble of projections from a single modelling group could therefore give a false sense of the certainty of the climate change signal for extreme precipitation events. There is a clear need to consider multiple ensembles of projections derived from multiple downscaling methodologies to account for this uncertainty, particularly if modelled outputs are to be used by decision makers. The estimates of changes per degree of global warming used in this study have all been derived from the SSP370 emissions scenario, as this was the highest emissions scenario shared by all the ensembles. Aerosols and land use changes from this scenario differ considerably from the other scenarios used for impact assessments (SSP126, 245, and 585), which may contribute to an underestimation of precipitation for some regions, particularly Asia 40 . However, forcing scenarios have been shown to be more influential for mean and moderately extreme precipitation compared to rare extremes, which are much more strongly linked to warming rates and not forcing agents 41 , 42 . Nonetheless, future work could compare the mean and moderately extreme precipitation from this study to other emissions scenarios (e.g. SSP585) to determine if these differences are significant. Projected changes to extreme precipitation are the result of thermodynamic and dynamic processes. Thermodynamic processes have been shown to lead to an increase to precipitation extremes in the order of 4 to 8% per degree warming globally 10 , while dynamic processes which influence the frequency and intensity of synoptic and subsynoptic features can increase or decrease this change but have high spatial variability and uncertainty. Robust reductions in the contribution of dynamic processes to extremes have been projected for subtropical regions, including parts of Australia 10 , which may explain the reduced rate of increase in moderately extreme precipitation shown in this study when compared to other regions 16 . It can be seen that the rarer events consistently have higher per degree changes, suggesting that they have larger contributions from dynamic processes. That is, the synoptic situation needs to be acting to enhance the thermodynamic effect in order to generate these rare extremes. In northern Australia, tropical cyclones are an important contributor of extreme precipitation, especially in the northwest where over 40% of extreme rainfall days (above the 99th percentile) are estimated to coincide with a tropical cyclone day 37 . Projections have pointed towards a reduction in the number of tropical cyclones impacting Australia, particularly the northwest, with less certain changes for the north central and northeast 43 , 44 . Reductions to the number of cyclones impacting northern Australia may help explain reductions to the moderate extremes noted for parts of northern Australia from some ensembles. However, some studies have suggested that there will be an increase in the intensity of these events, possibly contributing to increased precipitation extremes 45 . Extratropical cyclones are a major contributing factor for precipitation extremes along Eastern Australia 46 , 47 and will likely become less frequent in the future contributing to mean precipitation decreases. However, precipitation extremes associated with these events are shown to increase roughly in line with CC relationship 47 , 48 . Convective thunderstorms are important contributors of extreme precipitation across Australia 49 , they have been observed to increase in intensity over recent decades, and are expected to intensify further due to climate change 50 . However, the dynamically downscaled models used in this study are resolved at spatial scales between 10 and 20 km, which are not adequate to explicitly represent convective processes. Development of very high resolution models (< 4 km), which are able to explicitly represent convection is currently ongoing at a regional scale 34 , 51 , and may lead to improved simulation of precipitation extremes 52 . Further work is required to better understand the different drivers behind the projected changes to precipitation extremes from each of the ensembles considered. Our results show increases in extreme precipitation events are likely across Australia and its greater capital city regions, elevating the risk of flood events. Across Australia, flooding is already the costliest natural disaster 53 , impacting on all regions and population centres 54 , 55 . Within cities, these increases may be compounded by continual urban expansion, which increases runoff and exacerbates flooding 56 . Urban expansion and population growth also work to increase the exposure risk to flooding 57 , 58 . This trend is evident in the recent past, with a near doubling of the global urban area in floodplains impacted by 1 in 100 AEP events between 1985 and 2015 59 . Continual population growth and urban expansion may therefore work synergistically with climate change to exacerbate not only the magnitude of flooding events but also the population at risk and the cost of the potential damages. These global issues necessitate concerted adaptation and planning measures to mitigate future development within at-risk floodplains and to improve resilience. In some cases, increases to the magnitude of the largest flooding events may see a push for upstream dams to be increasingly used for flood mitigation instead of water supply 60 , which would necessitate dams to operate at lower maximum storages to accommodate larger flow volumes. Projected mean rainfall declines and the subsequent declines to the smaller more frequent flood events, which are important for water supply 61 could also reduce water security for some regions. Dam managers may find it increasingly hard to prioritise flood mitigation at the same time as water security as these two priorities become increasingly at odds with one another. Concurrently, more intense precipitation will likely exacerbate water quality issues by elevating erosion and nutrient runoff from agricultural lands 62 , which can have ecological impacts and cause drinking water supply issues. The projected changes to precipitation extremes highlighted in this study will therefore have a host of ramifications for flooding, the environment, dam management, water supply, and agriculture within Australia. To conclude, our analysis explores a large ensemble of CMIP6-based regional projections, including multiple GCM host models, regional models and RCM configurations to understand changes in regional extreme precipitation. It utilizes an innovative approach combining dynamical downscaling, generalised extreme value distribution and global warming level analysis to unravel the impacts of climate change in rare extreme precipitation events across greater capital cities, where approximately two thirds of the Australian population reside. The findings revealed globally relevant scientific insights and can inform decision making around topical issues such as flood management, urban water supply, urban planning, and agriculture. 4 Methods 4.1 Study Area This study evaluated changes to extreme precipitation across the Australian continent, which encompasses a range of climate regions, including arid, equatorial, savannah, subtropical, temperate, and tropical regions. We further examined the changes within greater capital city regions, where approximately two thirds of the Australian population live (Fig. 10 ). 4.2 Data We used a combined ensemble of high-resolution dynamically downscaled climate simulations for Australia generated by 4 Australian modelling groups. The 4 modelling groups applied 3 independent RCMs for dynamical downscaling, used in 5 different configurations. The GCMs selected for downscaling were based on different selection criteria 63 – 65 , which aimed to best represent the future spread in the climate change signal from the ensemble of CMIP6 models, while prioritising models which were statistically independent and better able to represent the Australian climate. In total, 19 different CMIP6 GCMs were chosen for downscaling, with some GCMs downscaled multiple times in different RCM configurations (Table 1 ). Table 1 Details of the different ensembles of climate considered in this study, consisting of 15 simulations from QldFCP-2 (oc denotes when downscaling involved ocean coupling), 7 simulations from CCAM-ACS, 7 simulations from BARPA-ACS, and 10 simulations from NARCliM2.0. An ensemble of 19 different CMIP6 GCMs was used for the downscaling. *For the GISS-E2-2-G model, r2i1p1f2 was downscaled, however, r1i1p1f1 was used for the host model comparison due to issues obtaining daily precipitation data for r2i1p1f2. GCMs RCMs CMIP6 Model Model full name Resolution Ensemble Member QldFCP-2 CCAM-ACS BARPA-ACS NARCliM2 ACCESS-ESM1.5 Australian Community Climate and Earth System Simulator, version 1.5 1.875 x 1.25° r6i1p1f1 CCAM CCAM_oc BARPA 2 x WRF r20i1p1f1 CCAM_oc r40i1p1f1 CCAM_oc ACCESS_CM2 Australian Community Climate and Earth System Simulator, version 2 1.875 x 1.25° r2i1p1f1 CCAM_oc r4i1p1f1 CCAM_oc BARPA CESM2 Community Earth System Model, version 2 1.25 x 0.9° r11i1p1f1 CCAM_oc BARPA CMCC-ESM2 Centro Euro-Mediterraneo sui Cambiamenti Climatici 1.25 x 0.9° r1i1p1f1 CCAM CCAM_oc BARPA CNRM-CM6-1-HR Centre National de Recherches Météorologiques Coupled Global Climate Model, version 6.1, high-resolution 0.5 x 0.5° r1i1p1f2 CCAM r1i1p1f2 CCAM_oc CNRM-ESM2 Centre National de Recherches Météorologiques Earth System Model, version 2 1 x 1° r1i1p1f2 CCAM_oc EC-Earth3 European Community Earth-System Model, version 3 0.8 x 0.8° r1i1p1f1 CCAM CCAM_oc BARPA EC-Earth3-Veg European Community Earth-System Model, version 3, Vegetation 0.8 x 0.8° r1i1p1f1 2 x WRF FGOALS-g3 Flexible Global Ocean-Atmosphere-Land System Model, grid point version 3 2.5 x 2.5° r4i1p1f1 CCAM GFDL-ESM4 Geophysical Fluid Dynamics Laboratory Earth System Model, version 4 1 x 1° r1i1p1f1 CCAM GISS-E2-2-G Goddard Institute for Space Studies Model E2.2G 2. x 2.5° r2i1p1f2 r1i1p1f2* CCAM MPI-ESM1-2-LR Max Planck Institute Earth System Model, version 1.2, low resolution 1.9 x 1.9° r9i1p1f1 CCAM MPI-ESM1-2-HR Max Planck Institute Earth System Model, version 1.2, high resolution 0.9 x 0.9° r1i1p1f1 BARPA 2 x WRF MRI-ESM2-0 Meteorological Research Institute Earth System Model, version 2.0 1.125 x 1.125° r1i1p1f1 CCAM NorESM2-MM Norwegian Earth System Model, version 2, 1 degree resolution 1 x 1° r1i1p1f1 CCAM CCAM_oc BARPA 2 x WRF r1i1p1f1 CCAM_oc UKESM1-0-LL United Kingdom Earth System Model, version 1.0 1.875 x 1.25° r1i1p1f2 2 x WRF QldFCP-2 and CCAM-ACS used the stretched grid Conformal Cubic Atmospheric Model CCAM; Thatcher, 2020) for downscaling 11 and 7 GCMs, respectively. QldFCP-2 downscaled to a 10 km spatial resolution over Australia, and used bias and variance corrected sea surface temperatures (SSTs) and sea ice, following the approach outlined by Hoffman et al. 66 . Five of the QldFCP-2 CCAM simulations were run using dynamic atmosphere-ocean coupling, while the rest were run in atmosphere-only mode (Table 1 ). CCAM-ACS downscaled to a 12.5 km spatial resolution and employed spectral nudging to constrain the model to follow the host GCM 67 , with all models run using dynamic atmosphere-ocean coupling. Both QldFPC-2 and CCAM-ACS used similar configurations for atmosphere, ocean, land-surface and aerosol parametrizations, and so the differences between them are mainly due to downscaling design (spectral nudging vs bias-corrected SSTs and sea ice), different GCMs, and differences in resolution. The Bureau of Meteorology Atmospheric Regional Projections for Australia (BARPA) is an RCM based on the UK Met Office Unified Model and Joint UK Land Environment Simulator (JULES) but configured for Australia and has been applied to downscale 7 GCMs to a 15 km spatial resolution 68 . Similar to CCAM-ACS, nudging was used to constrain the model to follow the host GCM 69 . NARCliM2.0 (New South Wales and Australian Regional Climate Modelling) employed two configurations of the Weather Research and Forecasting (WRF) model, each adopting different parameterisations of physics to downscale 5 GCMs. The two different sets of RCM parameterizations were selected based on their ability to simulate Australia’s recent climate and statistical independence from a larger set of 78 structurally different configurations. NARCliM2.0 provides simulations at a 20 km spatial resolution over Australasia. BARPA and NARCliM2.0 follow a limited area modelling approach, and as such these RCMs were forced at the lateral boundaries and used sea surface temperature from the host models. In total, a large ensemble consisting of 39 different regional models were considered in this study across the 4 modelling groups. Further details on the individual downscaling experiment designs can be found in 20 , 34 – 36 . Daily observed gridded precipitation data with a spatial resolution of 0.05° (approximately 5 km) were obtained from the Australian Gridded Climate Data Project (AGCD 70 ). Prior to analysis all datasets, including GCMs, RCMs, and observations were re-gridded to the same spatial resolution (i.e. 10 km) using distance weighting interpolation. We assessed the performance of host and downscaled models against the observational data using daily mean, 99th percentile precipitation (p99), and 99.7th percentile precipitation (p99.7). Regions with poor observational data quality during the comparison period were masked out of the model evaluation (Fig. 1 ). Model performance using these metrics was evaluated over Australia and over the eight greater capital city regions (Fig. 10 ) using the Mean Absolute Percentage Error (MAPE) and the Root Mean Square Error (RMSE) to quantify differences. 4.3 Extreme value analysis Extreme value analysis was applied to assess changes to the probability distribution of rare extreme events. Here these events represent the 1 in 10, 1 in 50, and 1 in 100 Annual Exceedance Probability (AEP) which approximately correspond to events with annual return intervals of 10 years, 50 years, and 100 years respectively. We sampled the daily timeseries of precipitation at each grid cell using the block maxima approach to derive annual maxima (AM) precipitation and then pooled together data from nearby cells using a 5x5 box centred on each grid cell to extend the data series used for the extreme event analysis 71 . For the analysis of capital cities, we chose to pool together all the AM data within each of the regions for the analysis. The Generalised Extreme Value (GEV) distribution was then fitted to the AM series using the L-moments method for parameter estimation. The GEV distribution is a generalised expression combining the Gumbel, Fréchet, and the Weibull distributions and is given by: $$\:G\left(x\right)=exp\left\{-{\left[1+\xi\:\left(\frac{x-\mu\:}{\sigma\:}\right)\right]}^{\frac{-1}{\xi\:}}\right\},\:for\:\left\{x:1+\xi\:\left(\frac{x-\mu\:}{\sigma\:}\right)>0\right\}$$ Where, µ , σ , and ξ are the location, scale, and shape parameters, respectively. Here, the location parameter is a measure of the central tendency and is loosely linked to the mean, the scale parameter is a measure of variance, and the shape parameter describes the tail behaviour. The shape parameter is particularly important for the estimation of rare extremes, as it describes the behaviour of the tail. When ξ ≥ 0 the distribution is unbounded with no upper limit, while when ξ < 0 the distribution is bounded by an upper limit 72 . Accurate estimation of the shape parameter necessitates many years of data to adequately fit, as it is susceptible to outliers 73 . We fit the GEV distribution to two 30-year periods representing the recent past (1981–2010) which we term the reference period, and the far future (2071–2100). However, as the data are pooled from nearby cells, we effectively increase the number of data points used in the analysis from 30 per grid cell to 750 71 . 4.4 Climate change assessment We examined the impacts of climate change on precipitation extremes by the end of the century (2071–2100) relative to the 1981–2010 reference period. We calculated the precipitation for the 1 in 10, 50, and 100 AEPs. We also evaluated changes to mean and moderately extreme (p99 and p99.7) precipitation to determine changes across a range of precipitation intensities. The 99.7th percentile precipitation was evaluated as this approximately corresponds to an event which would occur once per year. Calculations were applied at each individual grid cell for each of the climate models considered. The results for each of the projections were assessed individually and by model ensemble (i.e., 4 groups of RCMs and 1 group of GCMs), including a combined ensemble of all downscaled projections (CORDEX ensemble) assessed for the greater capital city regions. Model ensemble averages and medians were calculated, with calculations of the averages adopting a one model one vote rule. This weights the models according to the number of simulations per host model (Table 1 ) and results in an 11-model average for QldFCP-2, a 5-model average for NARCliM2.0, and a 7-model model average for CCAM-ACS and BARPA-ACS. For the CMIP6 GCMs we also weighted the two EC-Earth3 and MPI-ESM1-2 models to give a 14-model weighted ensemble. Resulting changes are presented as spatial maps over Australia using the ensemble average or median change and as boxplots for the eight greater capital city regions assessed using all model projections to better understand uncertainty. Averages were used for the analysis of mean, p99, and p99.7, while medians were used for the results of the GEV extremes to ensure they were not influenced by outliers. We present precipitation changes as a rate per degree of global temperature change (i.e., the global mean temperature, including both land and ocean regions). For consistency, we used the global temperature changes derived from the host GCMs as opposed to warming from the downscaled projections. This was implemented to ensure consistency between the different model ensembles and due to difficulty in deriving global warming levels from limited area RCMs such as BARPA and WRF. We calculated the projected changes per degree warming for the high emissions scenario (SSP370) only, as only this scenario and SSP126 were shared by all modelling groups, and as SSP370 would show a more sensitive response by the end of the century. Scaling rates were calculated for each model from the projected changes between the reference period (1981–2010) and the end of the century (2071–2100) as per the equation below. $$\:\frac{\%\:change}{degree\:warming}=\:\left[\left\{{\left(\frac{{P}_{f}}{{P}_{r}}\right)}^{\left(\frac{1}{{T}_{f}-{T}_{r}}\right)}\right\}-1\right]\times\:100$$ Where, P f is the precipitation in the future period, P r is the precipitation in the reference period, T f is the mean global temperature in the future period, and T r is the mean global temperature in the reference period. To determine where there is confidence in the scaling rate change over Australia, we adopt the signal-to-noise ratio to see where the climate change signal emerges over the ‘noise’ for each of the climate model ensembles considered 74 . Here, we consider the noise as the standard deviation from all the models in each ensemble. Stippling is shown on the ensemble mean and median change maps where the signal-to-noise ratio is greater than 1.0, as this is a commonly adopted threshold used in the literature 74 , 75 . To compare differences between the downscaled models and host models, we also compared the Probability Density Function (PDF) plot of the changes from these RCMs and GCMs. Here, the downscaled projections were also weighted evenly between modelling groups to avoid biasing the results from the larger downscaled ensembles. Declarations Data availability All data used in this study are publicly available. The downscaled climate projections which contribute to the CORDEX-CMIP6 Australasia domain can be accessed through the National Computer Infrastructure: https://nci.org.au/. The QldFCP-2 dataset is available at: https://dx.doi.org/10.25914/8fve-1910. The CCAM-ACS dataset is available at: https://dx.doi.org/10.25914/3r9s-pb86. The BARPA dataset is available at: https://dx.doi.org/10.25914/z1x6-dq28. The NARCliM2.0 dataset is available at: https://dx.doi.org/10.25914/3r9s-pb86. The CMIP6 global climate model data are available through the Earth System Grid Federation at: http://esgf.llnl.gov/. Gridded AGCD observations can be accessed at: https://dx.doi.org/10.25914/6009600b58196. Code availability All relevant codes used in this work are available upon request from the corresponding author. Competing interests All authors declare no financial or non-financial competing interests. Author Contribution R.E.: Writing – Original draft preparation, Conceptualization, Methodology, Formal analysis. J.S.: Conceptualization, Data Curation, Methodology. R.T.: Conceptualization, Methodology, Writing - Review & Editing. S.C.: Data Curation, Writing - Review & Editing. C.W.: Conceptualization, Methodology, Writing - Review & Editing. J.E.: Conceptualization, Methodology, Writing - Review & Editing. M.T.: Data Curation, Writing - Review & Editing. G.D.V.: Data Curation, Writing - Review & Editing. C.S.: Data Curation, Writing - Review & Editing Acknowledgement We acknowledge support by Lindsay Brebber from Information and Digital Science Delivery of the Department of Environment and Science for support with high performance computing and data storage. References Alexander, L. V. & Arblaster, J. M. Historical and projected trends in temperature and precipitation extremes in Australia in observations and CMIP5. Weather Clim. Extrem. 15, 34–56 (2017). Contractor, S., Donat, M. G. & Alexander, L. V. Intensification of the Daily Wet Day Rainfall Distribution Across Australia. Geophys. Res. Lett. 45, 8568–8576 (2018). Groisman, P. Y. et al. Trends in intense precipitation in the climate record. J. Clim. 18, 1326–1350 (2005). Jayaweera, L., Wasko, C., Nathan, R. & Johnson, F. Non-stationarity in extreme rainfalls across Australia. J. Hydrol. 624, 129872 (2023). Eccles, R., Zhang, H. & Hamilton, D. A review of the effects of climate change on riverine flooding in subtropical and tropical regions. J. Water Clim. Change 10, 687–707 (2019). Huang, X. et al. Global Projection of Flood Risk With a Bivariate Framework Under 1.5–3.0°C Warming Levels. Earths Future 12, e2023EF004312 (2024). Westra, S., Alexander, L. V. & Zwiers, F. W. Global Increasing Trends in Annual Maximum Daily Precipitation. (2013) doi: 10.1175/JCLI-D-12-00502.1 . Myhre, G. et al. Frequency of extreme precipitation increases extensively with event rareness under global warming. Sci. Rep. 9, 16063 (2019). Trenberth, K. E., Dai, A., Rasmussen, R. M. & Parsons, D. B. The Changing Character of Precipitation. Bull. Am. Meteorol. Soc. 84, (2003). Pfahl, S., O’Gorman, P. A. & Fischer, E. M. Understanding the regional pattern of projected future changes in extreme precipitation. Nat. Clim. Change 7, 423–427 (2017). Fowler, H. J. et al. Anthropogenic intensification of short-duration rainfall extremes. Nat. Rev. Earth Environ. 2, 107–122 (2021). O’Gorman, P. A. Precipitation Extremes Under Climate Change. Curr. Clim. Change Rep. 1, 49–59 (2015). Pendergrass, A. G. What precipitation is extreme? Science 360, 1072–1073 (2018). Grose, M. R. et al. Insights From CMIP6 for Australia’s Future Climate. Earths Future 8, e2019EF001469 (2020). Gründemann, G. J., van de Giesen, N., Brunner, L. & van der Ent, R. Rarest rainfall events will see the greatest relative increase in magnitude under future climate change. Commun. Earth Environ. 3, 1–9 (2022). Li, C. et al. Changes in Annual Extremes of Daily Temperature and Precipitation in CMIP6 Models. (2021) doi: 10.1175/JCLI-D-19-1013.1 . Abdelmoaty, H. M. & Papalexiou, S. M. Changes of Extreme Precipitation in CMIP6 Projections: Should We Use Stationary or Nonstationary Models? J. Clim. 36, 2999–3014 (2023). Reder, A., Raffa, M., Montesarchio, M. & Mercogliano, P. Performance evaluation of regional climate model simulations at different spatial and temporal scales over the complex orography area of the Alpine region. Nat. Hazards 102, 151–177 (2020). Boé, J. & Terray, L. Land–sea contrast, soil-atmosphere and cloud-temperature interactions: interplays and roles in future summer European climate change. Clim. Dyn. 42, 683–699 (2014). Chapman, S. et al. Evaluation of Dynamically Downscaled CMIP6-CCAM Models Over Australia. Earths Future 11, e2023EF003548 (2023). Grose, M. R. et al. The role of topography on projected rainfall change in mid-latitude mountain regions. Clim. Dyn. 53, 3675–3690 (2019). Tian, T. et al. Resolved complex coastlines and land–sea contrasts in a high-resolution regional climate model: a comparative study using prescribed and modelled SSTs. Tellus Dyn. Meteorol. Oceanogr. 65, 19951 (2013). Giorgi, F. Thirty years of regional climate modeling: Where are we and where are we going next? J. Geophys. Res. Atmospheres 124, 2018JD030094 (2019). Gutowski, W. J. et al. The Ongoing Need for High-Resolution Regional Climate Models: Process Understanding and Stakeholder Information. Bull. Am. Meteorol. Soc. 101, E664–E683 (2020). Zhang, X. et al. Indices for monitoring changes in extremes based on daily temperature and precipitation data. Wiley Interdiscip. Rev. Clim. Change 2, 851–870 (2011). Bao, J., Sherwood, S. C., Alexander, L. V. & Evans, J. P. Future increases in extreme precipitation exceed observed scaling rates. Nat. Clim. Change 7, 128–132 (2017). Chapman, S., Syktus, J., Trancoso, R., Toombs, N. & Eccles, R. Projected changes in mean climate and extremes from downscaled high-resolution CMIP6 simulations in Australia. Weather Clim. Extrem. 46, 100733 (2024). Ji, F. et al. Introducing NARCliM1.5: Evaluation and projection of climate extremes for southeast Australia. Weather Clim. Extrem. 38, 100526 (2022). Herold, N. et al. Projected changes in the frequency of climate extremes over southeast Australia. Environ. Res. Commun. 3, 011001 (2021). Mantegna, G. A., White, C. J., Remenyi, T. A., Corney, S. P. & Fox-Hughes, P. Simulating sub-daily Intensity-Frequency-Duration curves in Australia using a dynamical high-resolution regional climate model. J. Hydrol. 554, 277–291 (2017). Wasko, C., Guo, D., Ho, M., Nathan, R. & Vogel, E. Diverging projections for flood and rainfall frequency curves. J. Hydrol. 620, 129403 (2023). Trancoso, R. et al. Significantly wetter or drier future conditions for one to two thirds of the world’s population. Nat. Commun. 15, 483 (2024). Kendon, E. J., Rowell, D. P., Jones, R. G. & Buonomo, E. Robustness of future changes in local precipitation extremes. J. Clim. 21, 4280–4297 (2008). Di Virgilio, G. et al. Design, evaluation, and future projections of the NARCliM2.0 CORDEX-CMIP6 Australasia regional climate ensemble. Geosci. Model Dev. 18, 671–702 (2025). Howard, E. et al. Evaluation of multi-season convection-permitting atmosphere – mixed-layer ocean simulations of the Maritime Continent. Geosci. Model Dev. 17, 3815–3837 (2024). Schroeter, B. J. E., Ng, B., Takbash, A., Rafter, T. & Thatcher, M. A Comprehensive Evaluation of Mean and Extreme Climate for the Conformal Cubic Atmospheric Model (CCAM). (2024) doi: 10.1175/JAMC-D-24-0004.1 . Lavender, S. L. & Abbs, D. J. Trends in Australian rainfall: contribution of tropical cyclones and closed lows. Clim. Dyn. 40, 317–326 (2013). Wasko, C. et al. A systematic review of climate change science relevant to Australian design flood estimation. Hydrol. Earth Syst. Sci. Discuss. 1–48 (2023) doi: 10.5194/hess-2023-232 . Wasko, C., Lu, W. T. & Mehrotra, R. Relationship of extreme precipitation, dry-bulb temperature, and dew point temperature across Australia. Environ. Res. Lett. 13, 074031 (2018). Shiogama, H. et al. Important distinctiveness of SSP3–7.0 for use in impact assessments. Nat. Clim. Change 13, 1276–1278 (2023). Lin, L., Wang, Z., Xu, Y. & Fu, Q. Sensitivity of precipitation extremes to radiative forcing of greenhouse gases and aerosols. Geophys. Res. Lett. 43, 9860–9868 (2016). Lin, L., Wang, Z., Xu, Y., Fu, Q. & Dong, W. Larger Sensitivity of Precipitation Extremes to Aerosol Than Greenhouse Gas Forcing in CMIP5 Models. J. Geophys. Res. Atmospheres 123, 8062–8073 (2018). Bell, S. S. et al. Projections of southern hemisphere tropical cyclone track density using CMIP5 models. Clim. Dyn. 52, 6065–6079 (2019). Lavender, S. L. & Walsh, K. J. E. Dynamically downscaled simulations of Australian region tropical cyclones in current and future climates. Geophys. Res. Lett. 38, 1–6 (2011). Walsh, K. et al. Natural hazards in Australia: storms, wind and hail. Clim. Change 139, 55–67 (2016). Dowdy, A. J. et al. Review of Australian east coast low pressure systems and associated extremes. Clim. Dyn. 53, 4887–4910 (2019). Pepler, A. S. & Dowdy, A. J. Australia’s Future Extratropical Cyclones. J. Clim. 35, 7795–7810 (2022). Pepler, A. S. et al. Projected changes in east Australian midlatitude cyclones during the 21st century. Geophys. Res. Lett. 43, 334–340 (2016). Dowdy, A. J. Climatology of thunderstorms, convective rainfall and dry lightning environments in Australia. Clim. Dyn. 54, 3041–3052 (2020). Prein, A. F. et al. The future intensification of hourly precipitation extremes. Nat. Clim. Change 7, 48–52 (2017). Su, C.-H. et al. BARRA-C2: Development of the Kilometre-Scale Downscaled Atmospheric Reanalysis over Australia . (2024). Fosser, G. et al. Convection-permitting climate models offer more certain extreme rainfall projections. Npj Clim. Atmospheric Sci. 7, 1–10 (2024). Deloitte. Building Resilience to Natural Disasters in Our States and Territories . https://australianbusinessroundtable.com.au/assets/documents/ABR_building-resilience-in-our-states-and-territories.pdf (2017). Wasko, C., Sharma, A. & Pui, A. Linking temperature to catastrophe damages from hydrologic and meteorological extremes. J. Hydrol. 602, 126731 (2021). Wright, D. B., Bosma, C. D. & Lopez-Cantu, T. U.S. Hydrologic Design Standards Insufficient Due to Large Increases in Frequency of Rainfall Extremes. Geophys. Res. Lett. 46, 8144–8153 (2019). Hettiarachchi, S., Wasko, C. & Sharma, A. Increase in flood risk resulting from climate change in a developed urban watershed – the role of storm temporal patterns. Hydrol. Earth Syst. Sci. 22, 2041–2056 (2018). Strader, S. M. & Ashley, W. S. The Expanding Bull’s-Eye Effect. Weatherwise 68, 23–29 (2015). Wing, O. E. J. et al. Estimates of present and future flood risk in the conterminous United States. Environ. Res. Lett. 13, 034023 (2018). Andreadis, K. M. et al. Urbanizing the floodplain: global changes of imperviousness in flood-prone areas. Environ. Res. Lett. 17, 104024 (2022). Van den Honert, R. C. & McAneney, J. The 2011 Brisbane floods: causes, impacts and implications. Water 3, 1149–1173 (2011). Dykman, C., Sharma, A., Wasko, C. & Nathan, R. Can annual streamflow volumes be characterised by flood events alone? J. Hydrol. 617, 128884 (2023). Eccles, R., Zhang, H., Hamilton, D., Trancoso, R. & Syktus, J. Impacts of climate change on nutrient and sediment loads from a subtropical catchment. J. Environ. Manage. 345, 118738 (2023). Di Virgilio, G. et al. Selecting CMIP6 GCMs for CORDEX Dynamical Downscaling: Model Performance, Independence, and Climate Change Signals. Earths Future 10, 1–24 (2022). Grose, M. R. et al. A CMIP6-based multi-model downscaling ensemble to underpin climate change services in Australia. Clim. Serv. 30, 100368 (2023). Trancoso, R., Syktus, J., Toombs, N. & Chapman, S. Assessing and Selecting CMIP6 GCMs Ensemble Runs Based on Their Ability to Represent Historical Climate and Future Climate Change Signal . https://meetingorganizer.copernicus.org/EGU23/EGU23-11412.html (2023) doi: 10.5194/egusphere-egu23-11412 . Hoffmann, P., Katzfey, J. J., McGregor, J. L. & Thatcher, M. Bias and variance correction of sea surface temperatures used for dynamical downscaling. J. Geophys. Res. 121, 12,877 – 12,890 (2016). Thatcher, M. & McGregor, J. L. Using a Scale-Selective Filter for Dynamical Downscaling with the Conformal Cubic Atmospheric Model. Mon. Weather Rev. 137, 1742–1752 (2009). Su, C.-H. et al. BARPA: New Development of ACCESS-Based Regional Climate Modelling for Australian Climate Service . (Bureau of Meteorology, 2022). Stassen, C. et al. Development and Assessment of Regional Atmospheric Nudging in ACCESS . (Bureau of Meteorology Melbourne, 2023). Evans, A., Jones, D., Smalley, R. & Lellyett, S. An enhanced gridded rainfall analysis scheme for Australia. Aust. Bur. Meteorol. Melb. VIC Aust. 66, 55–67 (2020). Li, J., Evans, J., Johnson, F. & Sharma, A. A comparison of methods for estimating climate change impact on design rainfall using a high-resolution RCM. J. Hydrol. 547, 413–427 (2017). Coles, S. An Introduction to Statistical Modeling of Extreme Values . (Springer-Verlag, Bristol, UK, 2001). Papalexiou, S. M. & Koutsoyiannis, D. Battle of extreme value distributions: A global survey on extreme daily rainfall. Water Resour. Res. 49, 187–201 (2013). Hawkins, E. & Sutton, R. The potential to narrow uncertainty in projections of regional precipitation change. Clim. Dyn. 37, 407–418 (2011). Chapman, S., Syktus, J., Trancoso, R., Toombs, N. & Eccles, R. Projected Changes in Mean Climate and Extremes from Downscaled High-Resolution Cmip6 Simulations in Australia. SSRN Scholarly Paper at https://doi.org/10.2139/ssrn.4836517 (2024). Additional Declarations No competing interests reported. Supplementary Files PrecipitationExtremesSupplementaryMaterialsv5.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 19 Apr, 2025 Reviews received at journal 14 Apr, 2025 Reviews received at journal 10 Apr, 2025 Reviewers agreed at journal 03 Apr, 2025 Reviewers agreed at journal 01 Apr, 2025 Reviewers agreed at journal 01 Apr, 2025 Reviewers invited by journal 31 Mar, 2025 Editor assigned by journal 24 Mar, 2025 Submission checks completed at journal 24 Mar, 2025 First submitted to journal 24 Mar, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6291947","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":445021926,"identity":"ab7169ba-b86d-4289-950a-7ded06b5813c","order_by":0,"name":"Rohan Eccles","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEklEQVRIie3PMWuDQBTA8Xcc6HJw6xVL8hUsgm3pEb+Kh+BkIVPoaBAui+3sx+hHSLhVnAOdSqFTB8EltIH0ri00g4LZSrk/iG/wx/MB2Gx/MvLzdvXTOmbCORtHMACqvgg6gWAyhlyu7jfdHPiUFlh1fMEjcMVyC+9c5G7t95Hzukm8CtKLSjmplzWpyMlzcYUezJD1EsYy3yOgYh+TEN9KFQMTkqFSBQCDJPggcNCEdt21PES/hL4NkVBvWZst4CG51rdrAjs1ATawhdThDfETc0t4VjaJkOYWkacTh73Oe4lbBk/kbjalq+Kl3S1mEXWTzbbdc0Jp8thHvjv+AQdiACHNMD5NYH/C9zabzfbf+wT7ElFXeDcocQAAAABJRU5ErkJggg==","orcid":"","institution":"Queensland Treasury","correspondingAuthor":true,"prefix":"","firstName":"Rohan","middleName":"","lastName":"Eccles","suffix":""},{"id":445021927,"identity":"7f5e3474-93a2-4ff3-b67b-7887c69038c3","order_by":1,"name":"Jozef Syktus","email":"","orcid":"","institution":"The University of Queensland","correspondingAuthor":false,"prefix":"","firstName":"Jozef","middleName":"","lastName":"Syktus","suffix":""},{"id":445021928,"identity":"56d3a29f-2d2c-4f9e-b2be-a98ab48f9f54","order_by":2,"name":"Ralph Trancoso","email":"","orcid":"","institution":"Queensland Treasury","correspondingAuthor":false,"prefix":"","firstName":"Ralph","middleName":"","lastName":"Trancoso","suffix":""},{"id":445021929,"identity":"67206aef-c149-4518-bcff-1a8176aa8d09","order_by":3,"name":"Sarah Chapman","email":"","orcid":"","institution":"Queensland Treasury","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Chapman","suffix":""},{"id":445021930,"identity":"9cf083b5-3e77-4b9a-851b-35897b1aa0fc","order_by":4,"name":"Conrad Wasko","email":"","orcid":"","institution":"The University of Sydney","correspondingAuthor":false,"prefix":"","firstName":"Conrad","middleName":"","lastName":"Wasko","suffix":""},{"id":445021931,"identity":"a6cd9547-93f3-4f48-944f-a3a36cc14e48","order_by":5,"name":"Jason P. Evans","email":"","orcid":"","institution":"University of New South Wales","correspondingAuthor":false,"prefix":"","firstName":"Jason","middleName":"P.","lastName":"Evans","suffix":""},{"id":445021932,"identity":"8e5eccf0-d13f-47dc-a923-f395bd510525","order_by":6,"name":"Marcus Thatcher","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Marcus","middleName":"","lastName":"Thatcher","suffix":""},{"id":445021933,"identity":"17f772ba-7e38-4eee-9504-194e9da4140a","order_by":7,"name":"Giovanni Virgilio","email":"","orcid":"","institution":"University of New South Wales","correspondingAuthor":false,"prefix":"","firstName":"Giovanni","middleName":"","lastName":"Virgilio","suffix":""},{"id":445021934,"identity":"9d0af295-f16a-4514-a1e2-0cf1e7c8403a","order_by":8,"name":"Christian Stassen","email":"","orcid":"","institution":"Bureau of Meteorology","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Stassen","suffix":""}],"badges":[],"createdAt":"2025-03-24 05:53:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6291947/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6291947/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":81026659,"identity":"4d52a27b-8ef4-41d8-b5a2-28a35647daf0","added_by":"auto","created_at":"2025-04-21 10:41:11","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":774203,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eEnsemble average bias of daily mean, 99\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile, and 99.7\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile precipitation (model – AGCD) for the CMIP6 host models, QldFCP-2, CCAM-ACS, BARPA-ACS, and NARCliM2.0 ensembles over the 1981-2020 time period. Model comparisons are performed on a common 10 km grid and highlighted values denote the MAPE (black) and RMSE (grey). Areas where precipitation is poor in observations are masked out.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/e4930eb231aab2f051edee7a.png"},{"id":81026663,"identity":"18cdb4b4-567d-4330-ad8c-00dc14d205ae","added_by":"auto","created_at":"2025-04-21 10:41:11","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":289621,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eAnnual averages of daily mean, 99\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile, and 99.7\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile precipitation from observations and ensemble mean bias (model – AGCD) over the 1981-2020 time period for the greater capital city regions.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/42ad626c7c7c213e63fc2660.png"},{"id":81026665,"identity":"f48f34f4-d165-4880-a8f5-333cbf6b2264","added_by":"auto","created_at":"2025-04-21 10:41:11","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":529574,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eComparison of ensemble mean change to annual daily mean, 99\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile, and 99.7\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile precipitation per degree of global temperature change derived from SSP370. Highlighted values show the average change over Australia in percent, stippling shows where the signal-to-noise ratio \u0026gt; 1.0, and boxplots show the variability in the average change over Australia from each of the model ensembles.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/19f49ea04634d123db61ba5a.png"},{"id":81026668,"identity":"3e8ee58d-aa65-4fe7-bd72-3ae60be24e59","added_by":"auto","created_at":"2025-04-21 10:41:11","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":645359,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eComparison of ensemble median change to 1 in 10, 50, and 100 AEP daily precipitation per degree of global temperature change derived from SSP370. Highlighted values show the average changes over Australia in percent from the multi-model median, stippling shows where the signal-to-noise ratio \u0026gt; 1.0, and boxplots show the variability of the median change over Australia from each of the model ensembles.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/b0b9d8fab8a0c2102d8e697a.png"},{"id":81028608,"identity":"112e0869-7c93-42c3-897c-326551202c35","added_by":"auto","created_at":"2025-04-21 11:05:11","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":198394,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eProjected changes to the 99.7\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile (mean) and the 1 in 100 AEP (median) precipitation event for each of the individual climate models against projected regional warming at the end of the century (2071-2100). Models are grouped by ensemble according to colour, with the letter corresponding to the host model. Ensemble averages are shown using + symbols. Boxplots show the variability of all host and downscaled models.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/a0654e2a73abdeaeb5137046.png"},{"id":81027319,"identity":"614e0b30-ef03-4bb7-bc93-d2feec735331","added_by":"auto","created_at":"2025-04-21 10:49:11","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":256765,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eChanges to daily mean, 99\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile, and 99.7\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e percentile precipitation per degree of global temperature change derived from SSP370. Boxplots show variability in projected changes within greater capital city regions from each of the model ensembles.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/d26505733a6e3455b3c936e3.png"},{"id":81026671,"identity":"d06621cb-201c-473d-a0e8-0ba0041dc28f","added_by":"auto","created_at":"2025-04-21 10:41:11","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":241240,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eChanges to 1 in 10, 50, and 100 AEP daily precipitation per degree of global temperature change derived from SSP370. Boxplots show variability in projected changes within greater capital city regions from each of the model ensembles.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/bd85d02b392e4b160e831458.png"},{"id":81027318,"identity":"a017868c-f550-4bea-8f48-0178aae7ceb9","added_by":"auto","created_at":"2025-04-21 10:49:11","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":553071,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eProbability density function plot of change to daily 1 in 10, 50, and 100 AEP daily precipitation per degree of global temperature change derived from SSP370. Coloured bars show median change for each of the individual downscaled ensembles, while dotted vertical lines show the host and downscaled model ensemble averages (each modelling group weighted evenly). Bold blue and red text show projected changes per degree warming from the host and downscaled ensembles, respectively.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/a5f70a9270a048065d179439.png"},{"id":81028609,"identity":"28500a44-9072-4b25-aa6d-7ddeafd28e90","added_by":"auto","created_at":"2025-04-21 11:05:11","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":324665,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSummary of projected changes to mean and extreme precipitation from downscaled and host models for greater capital city regions: (a) Darwin, (b) Brisbane, (c) Sydney, (d) Australian Capital Territory, (e) Melbourne, (f) Hobart, (g) Adelaide, and (h) Perth.\u003c/em\u003e \u003cem\u003eError bars show the 10th and 90th percentile of changes. \u0026nbsp;Background shows the mean precipitation change per degree global warming across Australia from the full ensemble of all downscaled projections (each modelling group weighted evenly).\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/007455197778041e92efcabd.png"},{"id":81026693,"identity":"b10ab7e7-dac6-44ed-b41b-3197609815dc","added_by":"auto","created_at":"2025-04-21 10:41:12","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":231510,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eExtent of study area and location of Australian greater capital city areas adopted in this study compared to major climate regions.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image10.png","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/0f7fd2334f5121cdfa3a9285.png"},{"id":81029947,"identity":"da644e24-859e-4abe-90e6-334747e8b1e5","added_by":"auto","created_at":"2025-04-21 11:13:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5001800,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/381241df-ff16-4e6d-88b8-40aa3f9e6799.pdf"},{"id":81028242,"identity":"ef57cf73-cc56-4799-9077-86c85d51c215","added_by":"auto","created_at":"2025-04-21 10:57:11","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1402222,"visible":true,"origin":"","legend":"","description":"","filename":"PrecipitationExtremesSupplementaryMaterialsv5.docx","url":"https://assets-eu.researchsquare.com/files/rs-6291947/v1/9fe50de11a24c92138d7772e.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Substantial increases in future precipitation extremes – insights from a large ensemble of downscaled CMIP6 models","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eThe intensity and frequency of extreme precipitation events is widely expected to increase as a result of climate change\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, which will have significant implications for future flooding\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Globally, observations show an increase in annual maximum daily precipitation between 5.9% and 7.7% per degree of warming\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Even larger increases have been reported for the 99th (11% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and 99.97th (13% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) percentile precipitation\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Observed changes are a result of thermodynamic and dynamic processes, both of which are influenced by a warming atmosphere\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Thermodynamic processes such as the Clausius-Clapeyron (CC) relationship, which increases the water-holding capacity of the atmosphere by 6\u0026ndash;7% per degree of warming\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e lead to relatively spatially homogeneous changes in precipitation intensity on the order of 4\u0026ndash;8% per degree of warming\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Daily precipitation extremes have been shown to be intensifying at a rate approximately equal to the CC relationship\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Dynamic processes substantially modify the regional response to climate warming, through changes to the frequency and intensity of synoptic and subsynoptic phenomena, including tropical and extratropical cyclones\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Increased warming also increases atmospheric stability, which weakens circulation, potentially reducing the intensity of precipitation extremes. On the other hand, latent heat release results in atmospheric instability that can strengthen storms and precipitation extremes, with the most extreme events seeing the largest increases\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. The overall impact of climate change on precipitation intensity is therefore highly complex and regionally dependent.\u003c/p\u003e \u003cp\u003eClimate models are currently the best physically based approaches available to understand future precipitation processes, characteristics, and impacts. Global Climate Models (GCMs) from the latest Coupled Model Intercomparison Project Phase 6 (CMIP6) have shown increases to precipitation extremes in-line with the CC relationship, with smaller increases noted for more moderate events compared to rarer events\u003csup\u003e\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Abdelmoaty \u0026amp; Papalexiou\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e showed historic 1 in 100 Annual Exceedance Probability (AEP) precipitation events would occur approximately every 50 and 70 years for the Northern and Southern Hemispheres, respectively by mid-century (2035\u0026ndash;2067). Using 25 GCMs, Gr\u0026uuml;ndemann et al.\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e found the intensity of 1 in 100 AEP precipitation events increased by between 13.5% and 38.3%, depending on the emissions scenario, with smaller increases for more frequent events. An ensemble of CMIP6 GCMs projected increases to 1 in 10 and 1 in 50 AEP daily and 5-daily precipitation at rates approximately equal to the CC relationship (~\u0026thinsp;7% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003csup\u003e16\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eGCMs are, however, applied at relatively coarse resolutions (~\u0026thinsp;150 km) and have difficulty adequately representing precipitation patterns over complex terrain and resolving extreme precipitation processes\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. These coarse resolution models are not always suited to providing reliable regionally specific information required to support adaptation and decision-making at regional scales. To overcome these issues, Regional Climate Models (RCMs) dynamically downscale GCMs to a finer resolution (typically 10 to 50 km) in order to better represent small-scale features and processes. RCMs have been shown to have improved skill in representing patterns of local precipitation and the impacts of topography, coasts, and land use changes compared to GCMs\u003csup\u003e\u003cspan additionalcitationids=\"CR20 CR21\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e and may also be more skilful in simulating extremes\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn Australia, most studies to date using RCMs have focused on percentiles or used extreme indices defined by the Expert Team on Climate Change Detection and Indices\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, such as the RX1day or the 99th percentile\u003csup\u003e\u003cspan additionalcitationids=\"CR27\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. While these moderate extreme events are important, most flood impacts are associated with rarer extremes, which are less well understood. Herold et al.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e projected 1 in 20 AEP daily precipitation would occur 1\u0026ndash;2 times more frequently in capital cities across southeast Australia using projections from 4 downscaled CMIP3 GCMs, though there was low model agreement for these changes. Mantegna et al.\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e showed daily intensities could increase by over 15% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the most extreme events (1 in 100 AEP) over Tasmania using 5 downscaled GCMs, with smaller increases for more moderate events. Similarly, Wasko, Guo et al.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e found larger increases for more extreme daily events (up to the 1 in 50 AEP) compared to less extreme events across Australia using 4 statistically downscaled GCMs, with larger increases for the north and smaller increases for the southwest.\u003c/p\u003e \u003cp\u003eThere are significant uncertainties associated with projections of precipitation from climate models\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, which necessitates large ensembles to better account for this uncertainty, especially when evaluating extremes\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. However, studies on rare extremes to date have largely adopted limited ensemble sizes and used earlier climate projections, which may underestimate the uncertainty. There is a need to determine how climate change influences precipitation extremes using a large ensemble of the latest downscaled CMIP6 projections, particularly for the rarest events.\u003c/p\u003e \u003cp\u003eThe focus of this paper is therefore to examine the impacts of climate change on precipitation extremes using a large ensemble of downscaled CMIP6 projections. For this purpose, we use downscaled ensembles from 4 modelling groups under a high emission Shared Socioeconomic Pathway (SSP370). This large ensemble forms the core of Australia\u0026rsquo;s new national projections, contributing to the CORDEX-CMIP6 Australasia domain\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan additionalcitationids=\"CR35\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e, and underpins Australia\u0026rsquo;s climate services and adaptation planning. This large ensemble consists of 19 different host GCMs downscaled using 3 distinct regional climate models in 5 configurations, for a total of 39 different downscaled simulations (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e; more details in methods section). This represents the largest ensemble of high-resolution climate projections used to assess changes to precipitation in Australia, allowing for a more robust assessment of the uncertainty and the spatial variability of likely changes. The inclusion of multiple different downscaled ensembles also allows for an assessment of the uncertainty introduced from the downscaling approach, which has not previously been quantified for rare precipitation extremes. We focus on greater capital cities as this is where two thirds of the Australian population lives and thus where the impacts are likely to be most pronounced. The aims of this paper are threefold:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eto assess the performance of the CMIP6 GCMs and different RCM ensembles in simulating mean and moderately extreme precipitation (i.e., 99th and 99.7th percentiles).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eto investigate changes and uncertainty in mean, moderately extreme (99th and 99.7th percentiles), and rare extreme precipitation (1 in 10, 50, and 100 AEP) per degree of global warming based on a large ensemble of regional models.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eto investigate how extreme precipitation varies spatially across the continent and identify regions with more pronounced extreme precipitation increases.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e"},{"header":"2 Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Evaluation of simulated daily mean and moderate extreme precipitation\u003c/h2\u003e \u003cp\u003eWe assess the performance of the host GCMs and the 4 downscaled ensembles in simulating daily mean and moderately extreme precipitation across Australia by comparing to observation data. The ensemble of CMIP6 GCMs performed reasonably well in representing mean and moderate extreme precipitation across Australia, with lower average errors over Australia than some of the downscaled ensembles (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The performance of the CMIP6 ensemble was, however, lacking over coastal and mountainous regions, with a consistent dry bias, particularly for the moderate extremes (p99 and p99.7). This is evident when evaluating biases for the greater capital city regions, which are all situated near to the coast and/or exhibit complex terrain. In these regions, where the majority of the population resides, the CMIP6 GCMs were shown to perform significantly worse in representing moderate extremes compared to the downscaled ensembles, though performance was comparable when evaluating mean precipitation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Of the downscaled ensembles, NARCliM2.0 and QldFCP-2 had notable dry biases over northern and western Australia, with smaller biases over southeast Australia (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). By contrast, CCAM-ACS was noted to have an opposing wet bias that was greatest in the interior, including for southeast Australia. BARPA-ACS was the best performing ensemble over Australia as a whole for both mean and moderately extreme precipitation, with minimal MAPE and RMSE values.\u003c/p\u003e \u003cp\u003eWithin the greater capital city regions, all downscaled ensembles reported lower biases for p99.7 compared to mean precipitation and p99, except for CCAM-ACS, which had lower biases for mean precipitation (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e to S3). For mean precipitation, CCAM-ACS had the lowest bias of any ensemble in the capital cities (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). QldFCP-2 and NARCliM2.0 both showed a clear dry bias over most capital cities for mean precipitation, whereas biases for CCAM-ACS and BARPA-ACS varied according to the city. For p99 and p99.7, the performance of all downscaled ensembles varied according to location (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). For p99.7, CCAM-ACS tended to overestimate precipitation intensity in capital cities, whereas NARCliM2.0 tended to underestimate intensity, though biases for NARCliM2.0 were relatively small. In contrast to the mean precipitation, QldFCP-2 also tended to have low biases for p99.7 except for in Perth and Hobart where the models showed a clear dry bias. BARPA-ACS also had good agreement except for in Brisbane and Adelaide where a wet bias was noted (Table S3). These biases are, however, in almost all cases an improvement when compared to the CMIP6 host models for the capital cities, suggesting that the downscaled models from these ensembles may be better suited to represent precipitation extremes in these regions compared to the host models.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Projected changes to mean, moderate and rare extreme precipitation\u003c/h2\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.2.1 Changes across Australia\u003c/h2\u003e \u003cp\u003eConsiderable spatial variability was evident in the projections of future mean and moderately extreme precipitation over Australia (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Spatial patterns of change varied according to the ensemble applied, with a consensus towards wetting or drying evident in a few regions only. A clear drying pattern for mean precipitation was evident in southwest Western Australia and southern Victoria and Tasmania to a lesser extent. This, however, did not necessarily lead to clear decreases for moderately extreme precipitation over these regions. In other regions, a drying or wetting signal was evident according to the signal-to-noise ratio for some ensembles but not for others. For example, the QldFCP-2 ensemble shows widespread agreement for decreased mean and p99 along coastal regions of north-eastern Australia, while the BARPA-ACS ensemble shows widespread agreement for increases over northern Australia, neither of which are reflected in the outputs for the other modelling groups. Similarly, over southeast Australia all model ensembles except for NARCliM2.0 show widespread agreement for increased p99.7. These results indicate that relying on a single ensemble of downscaled projections may give a false sense of the certainty of precipitation changes.\u003c/p\u003e \u003cp\u003eAll model ensembles show precipitation to vary according to the metric assessed, with the smallest (greatest) increases (decreases) seen for mean precipitation, with greater increases noted for more intense events compared to less intense events. Ensemble mean precipitation was projected to change by between \u0026minus;\u0026thinsp;2.5% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 2.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, while p99.7 changed by between \u0026minus;\u0026thinsp;0.1% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 4% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, depending on the ensemble used (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). These changes are greater when evaluating the response of rare extremes to climate change (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), where ensemble median changes ranged from 2.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 6.5% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the 1 in 10 AEP event and from 4.5% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 10.1% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the 1 in 100 AEP event. Both NARCliM2.0 and QldFCP-2 projected smaller increases to rare precipitation extremes over Australia than the CMIP6 host models (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). By contrast, CCAM-ACS and BARPA-ACS projected increases greater than the CMIP6 host models. The increases from most ensembles tended to be greatest over northern Australia, especially from the BARPA-ACS ensemble. Despite this inter-ensemble agreement, there were few areas in any ensemble which showed agreement for the 1 in 100 AEP according to the signal-to-noise ratio, due to the variability of future extremes. Corresponding changes to the spatial patterns of the GEV fitted parameter values are presented in the Supplementary Materials (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOver Australia, the intra-model variability tended to be greater within the NARCliM2.0 ensemble than the other modelling groups, despite sampling the least GCMs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). This is especially evident for p99.7 and appears to be due to the very wet model (EC-Earth3-veg) and very dry (ACCESS-ESM) becoming more wet and dry following downscaling (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). By contrast, the other downscaling methods tended to bring these outlier models in towards the mean, with some exceptions, including EC-Earth3 downscaled with BARPA-ACS which showed the largest increases of any model assessed for both the 99.7th percentile and the 1 in 100 AEP event (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The QldFCP-2 ensemble tended to have the least variability in the projections, followed by the CCAM-ACS. As both ensembles share the same model for downscaling, this could relate to the CCAM model effectively dampening the magnitude of changes from outlier models.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.2.2 Changes within Capital Cities\u003c/h2\u003e \u003cp\u003eThere was widespread agreement from all ensembles for a decrease to mean precipitation (-11.2% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to -8.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), p99 (-5.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to -4.2% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), and p99.7 (-3.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to -1.5% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) in Perth (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e) with smaller decreases for more intense events compared to the mean. In other capital cities there was less consensus in the magnitude or sign of change, though generally decreases to mean precipitation were projected from most ensembles in Adelaide and Melbourne, with only CCAM-ACS tending towards a slight increase. Most capital cities tended towards an increase to the more intense events (p99.7) compared to the mean, with the largest increases typically noted for Darwin (-2.3% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 8.3% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). In comparison to the projections for the means and moderate extremes, rare extremes tended towards an increase for almost all ensembles in all greater capital cities. Darwin noted the largest ensemble median increases in the 1 in 100 AEP event (8.3% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 18.2% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), with the greatest increases reported from the BARPA ensemble (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHere, the BARPA ensemble appears to show greater projected changes to extremes compared to the other CORDEX ensembles (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Similarly, NARCliM2.0 appears to show lower projected changes in Melbourne, while QldFCP-2 shows smaller changes in Brisbane when compared to the other CORDEX ensembles. In general, there were considerable differences in the projections across all capital cities, particularly those in northern and eastern Australia (Brisbane, Darwin, Sydney, ACT, and Melbourne). This variability is evident when comparing the distribution of the combined downscaled CORDEX ensemble to the individual downscaled ensembles (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), highlighting the uncertainty associated with projections of precipitation and the need to adopt large ensembles using different downscaling methodologies to quantify the uncertainty. Generally, the combined downscaled ensemble appears to have a similar spread of projected changes to the host model ensemble, except in those regions discussed above (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eand Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe median projection of change from the host model ensemble closely matched that from the weighted downscaled ensemble (each modelling group weighted evenly; Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). This was, however, not true for Adelaide where the host models showed 1 in 100 AEP was projected to increase by 9% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in Adelaide compared to 4% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e from the weighted downscaled ensemble or Brisbane where the host models showed a 3.7% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e compared to 7.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e from the weighted downscaled ensemble. Both groups projected the largest increases to the 1 in 100 AEP for Darwin (12.2% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 11.9% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the host and downscaled ensembles, respectively), which is considerably larger than what would be expected from the CC relationship (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3 Discussion","content":"\u003cp\u003eOur study shows downscaling consistently reduces biases of moderately extreme precipitation (p99 and p99.7) over capital city regions in Australia when compared to the host CMIP6 GCMs (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Table S2 to S3). In contrast, the host models performed comparably well for mean precipitation within the capital city regions and for all mean and moderately extreme precipitation when assessed across Australia as a whole. These differences relate to the coarse model resolution of the GCMs, which are not able to represent precipitation patterns and extremes over regions with complex terrain, land use changes, or coastal gradients when compared to finer resolved RCMs\u003csup\u003e\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. It is, however, within these complex regions where the majority of the Australian population resides and where the performance differences between host and downscaled models are greatest. RCMs therefore appear to be better suited to provide information on climate hazards within these populated areas, particularly for extremes.\u003c/p\u003e \u003cp\u003eThere were considerable differences in the spatial extent of biases from the different downscaled modelling ensembles. Both NARCliM2.0 and QldFCP-2 had notable dry biases over Australia and the capital city regions, while CCAM-ACS and BARPA-ACS had notable wet biases, though to a lesser extent. The dry biases from the QldFCP-2 ensemble are particularly evident over northern Australia and may relate to a misrepresentation of the number and intensity of low-pressure systems and cyclones, which are major contributors of precipitation, especially extremes in these regions\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. NARCliM2.0 shows similar dry biases over northern Australia and within the greater capital city regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) for mean and p99, but much less bias for p99.7. Di Virgilio et al.\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e posited that the dry bias over northern Australia could relate to issues in capturing the Australian monsoon. They also showed that there was less bias in mean precipitation over southeast Australia from NARCliM2.0 compared to previous downscaled models from CMIP5 (NARCliM1.5) and CMIP3 (NARCliM1.0). By contrast, the wet bias present in the CCAM-ACS ensemble has been shown to relate to an overestimation of extreme precipitation, while low precipitation events were underestimated from this ensemble\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. These biases are believed to relate to the parameterisation of CCAM, with improved schemes currently being tested to resolve these issues. Similarly, the wet bias from the BARPA ensemble has been shown to relate to a general overestimation of precipitation extremes, and inclusion of too many low intensity wet days. It is interesting to note the different sign of the biases between the CCAM-ACS and the QldFCP-2 as both ensembles make use of the CCAM model for downscaling. These differences likely relate to the downscaling approach adopted, particularly in regard to sea surface temperatures, which have a significant influence on precipitation. Here, QldFCP-2 elected to bias correct sea-surface temperatures from the host models, whereas CCAM-ACS adopted a nudging approach for the atmosphere and sea-surface temperatures, leading to diverging simulations of precipitation from the same host model.\u003c/p\u003e \u003cp\u003eExtreme precipitation was projected to increase across Australia and all capital cities, with greater increases seen for rarer extremes compared to more moderate extremes (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e), in line with previous studies\u003csup\u003e\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. The magnitude of the projected changes were, however, dependent on the combination of the model ensemble and region considered. We found that changes for the 99.7th percentile precipitation ranged from between \u0026minus;\u0026thinsp;0.1% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 4.0% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), while changes to the 1 in 100 AEP ranged from between 4.5% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 10.1% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e across Australia (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), which are the result of increases to all GEV parameters (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e and Figure S2). These projected changes are broadly in line with the findings of Wasko, Westra, et al.\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, who suggested a scaling rate of 8% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for Australia from a meta-analysis of available studies of observations and projections. To date, available studies on projections have made it difficult to ascertain if there are any geographic differences in these scaling rates\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, as the majority of the literature to date has focused on the populated southeast Australia\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eUsing the largest ensemble of projections to date, we show that there does appear to be large-scale geographic differences in projected precipitation extremes, with greater increases generally projected for northern Australia compared to southern Australia. Similar findings have been noted when assessing the observational records\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e and from a recent study based on 4 downscaled CMIP5 GCMs\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. While this north-south difference is generally evident across the ensembles, it is important to note that there are widespread regional variations from the different ensembles for all precipitation intensities, except for southwest Australia where a strong signal of decreasing mean precipitation is shown for all ensembles (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Decreased mean precipitation, however, did not necessarily translate to a decrease in moderate or rare extremes, with generally consistent increases to rare extremes still projected for Perth despite the declines to mean precipitation (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn other regions, a drying or wetting signal is evident according to the signal-to-noise ratio for some ensembles but not for others. For example, the QldFCP-2 ensemble reported significant decreases to mean and 99th percentile precipitation for coastal regions of north-eastern Australia, whereas BARPA-ACS noted significant increases over northern Australia, neither of which are reflected in other modelling group outputs. Relying on a single ensemble of projections from a single modelling group could therefore give a false sense of the certainty of the climate change signal for extreme precipitation events. There is a clear need to consider multiple ensembles of projections derived from multiple downscaling methodologies to account for this uncertainty, particularly if modelled outputs are to be used by decision makers.\u003c/p\u003e \u003cp\u003eThe estimates of changes per degree of global warming used in this study have all been derived from the SSP370 emissions scenario, as this was the highest emissions scenario shared by all the ensembles. Aerosols and land use changes from this scenario differ considerably from the other scenarios used for impact assessments (SSP126, 245, and 585), which may contribute to an underestimation of precipitation for some regions, particularly Asia\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. However, forcing scenarios have been shown to be more influential for mean and moderately extreme precipitation compared to rare extremes, which are much more strongly linked to warming rates and not forcing agents\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Nonetheless, future work could compare the mean and moderately extreme precipitation from this study to other emissions scenarios (e.g. SSP585) to determine if these differences are significant.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eProjected changes to extreme precipitation are the result of thermodynamic and dynamic processes. Thermodynamic processes have been shown to lead to an increase to precipitation extremes in the order of 4 to 8% per degree warming globally\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, while dynamic processes which influence the frequency and intensity of synoptic and subsynoptic features can increase or decrease this change but have high spatial variability and uncertainty. Robust reductions in the contribution of dynamic processes to extremes have been projected for subtropical regions, including parts of Australia\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, which may explain the reduced rate of increase in moderately extreme precipitation shown in this study when compared to other regions\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. It can be seen that the rarer events consistently have higher per degree changes, suggesting that they have larger contributions from dynamic processes. That is, the synoptic situation needs to be acting to enhance the thermodynamic effect in order to generate these rare extremes.\u003c/p\u003e \u003cp\u003eIn northern Australia, tropical cyclones are an important contributor of extreme precipitation, especially in the northwest where over 40% of extreme rainfall days (above the 99th percentile) are estimated to coincide with a tropical cyclone day\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Projections have pointed towards a reduction in the number of tropical cyclones impacting Australia, particularly the northwest, with less certain changes for the north central and northeast\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Reductions to the number of cyclones impacting northern Australia may help explain reductions to the moderate extremes noted for parts of northern Australia from some ensembles. However, some studies have suggested that there will be an increase in the intensity of these events, possibly contributing to increased precipitation extremes\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Extratropical cyclones are a major contributing factor for precipitation extremes along Eastern Australia\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e,\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e and will likely become less frequent in the future contributing to mean precipitation decreases. However, precipitation extremes associated with these events are shown to increase roughly in line with CC relationship\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eConvective thunderstorms are important contributors of extreme precipitation across Australia\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, they have been observed to increase in intensity over recent decades, and are expected to intensify further due to climate change\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. However, the dynamically downscaled models used in this study are resolved at spatial scales between 10 and 20 km, which are not adequate to explicitly represent convective processes. Development of very high resolution models (\u0026lt;\u0026thinsp;4 km), which are able to explicitly represent convection is currently ongoing at a regional scale\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e, and may lead to improved simulation of precipitation extremes\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Further work is required to better understand the different drivers behind the projected changes to precipitation extremes from each of the ensembles considered.\u003c/p\u003e \u003cp\u003eOur results show increases in extreme precipitation events are likely across Australia and its greater capital city regions, elevating the risk of flood events. Across Australia, flooding is already the costliest natural disaster\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e, impacting on all regions and population centres\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e,\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Within cities, these increases may be compounded by continual urban expansion, which increases runoff and exacerbates flooding\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. Urban expansion and population growth also work to increase the exposure risk to flooding\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. This trend is evident in the recent past, with a near doubling of the global urban area in floodplains impacted by 1 in 100 AEP events between 1985 and 2015\u003csup\u003e59\u003c/sup\u003e. Continual population growth and urban expansion may therefore work synergistically with climate change to exacerbate not only the magnitude of flooding events but also the population at risk and the cost of the potential damages. These global issues necessitate concerted adaptation and planning measures to mitigate future development within at-risk floodplains and to improve resilience.\u003c/p\u003e \u003cp\u003eIn some cases, increases to the magnitude of the largest flooding events may see a push for upstream dams to be increasingly used for flood mitigation instead of water supply\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e, which would necessitate dams to operate at lower maximum storages to accommodate larger flow volumes. Projected mean rainfall declines and the subsequent declines to the smaller more frequent flood events, which are important for water supply\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e could also reduce water security for some regions. Dam managers may find it increasingly hard to prioritise flood mitigation at the same time as water security as these two priorities become increasingly at odds with one another. Concurrently, more intense precipitation will likely exacerbate water quality issues by elevating erosion and nutrient runoff from agricultural lands\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e, which can have ecological impacts and cause drinking water supply issues. The projected changes to precipitation extremes highlighted in this study will therefore have a host of ramifications for flooding, the environment, dam management, water supply, and agriculture within Australia.\u003c/p\u003e \u003cp\u003eTo conclude, our analysis explores a large ensemble of CMIP6-based regional projections, including multiple GCM host models, regional models and RCM configurations to understand changes in regional extreme precipitation. It utilizes an innovative approach combining dynamical downscaling, generalised extreme value distribution and global warming level analysis to unravel the impacts of climate change in rare extreme precipitation events across greater capital cities, where approximately two thirds of the Australian population reside. The findings revealed globally relevant scientific insights and can inform decision making around topical issues such as flood management, urban water supply, urban planning, and agriculture.\u003c/p\u003e"},{"header":"4 Methods","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Study Area\u003c/h2\u003e \u003cp\u003eThis study evaluated changes to extreme precipitation across the Australian continent, which encompasses a range of climate regions, including arid, equatorial, savannah, subtropical, temperate, and tropical regions. We further examined the changes within greater capital city regions, where approximately two thirds of the Australian population live (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Data\u003c/h2\u003e \u003cp\u003eWe used a combined ensemble of high-resolution dynamically downscaled climate simulations for Australia generated by 4 Australian modelling groups. The 4 modelling groups applied 3 independent RCMs for dynamical downscaling, used in 5 different configurations. The GCMs selected for downscaling were based on different selection criteria\u003csup\u003e\u003cspan additionalcitationids=\"CR64\" citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e, which aimed to best represent the future spread in the climate change signal from the ensemble of CMIP6 models, while prioritising models which were statistically independent and better able to represent the Australian climate. In total, 19 different CMIP6 GCMs were chosen for downscaling, with some GCMs downscaled multiple times in different RCM configurations (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDetails of the different ensembles of climate considered in this study, consisting of 15 simulations from QldFCP-2 (oc denotes when downscaling involved ocean coupling), 7 simulations from CCAM-ACS, 7 simulations from BARPA-ACS, and 10 simulations from NARCliM2.0. An ensemble of 19 different CMIP6 GCMs was used for the downscaling. *For the GISS-E2-2-G model, r2i1p1f2 was downscaled, however, r1i1p1f1 was used for the host model comparison due to issues obtaining daily precipitation data for r2i1p1f2.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eGCMs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c8\" namest=\"c5\"\u003e \u003cp\u003eRCMs\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCMIP6 Model\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModel full name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eResolution\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnsemble Member\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eQldFCP-2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM-ACS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA-ACS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eNARCliM2\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eACCESS-ESM1.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eAustralian Community Climate and Earth System Simulator, version 1.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e1.875 x 1.25\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er6i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2 x WRF\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er20i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er40i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eACCESS_CM2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eAustralian Community Climate and Earth System Simulator, version 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.875 x 1.25\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er2i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er4i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCESM2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCommunity Earth System Model, version 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.25 x 0.9\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er11i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCMCC-ESM2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCentro Euro-Mediterraneo sui Cambiamenti Climatici\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.25 x 0.9\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eCNRM-CM6-1-HR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eCentre National de Recherches M\u0026eacute;t\u0026eacute;orologiques Coupled Global Climate Model, version 6.1, high-resolution\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.5 x 0.5\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCNRM-ESM2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCentre National de Recherches M\u0026eacute;t\u0026eacute;orologiques Earth System Model, version 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 x 1\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEC-Earth3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEuropean Community Earth-System Model, version 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.8 x 0.8\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEC-Earth3-Veg\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEuropean Community Earth-System Model, version 3, Vegetation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.8 x 0.8\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2 x WRF\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFGOALS-g3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFlexible Global Ocean-Atmosphere-Land System Model, grid point version 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.5 x 2.5\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er4i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGFDL-ESM4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGeophysical Fluid Dynamics Laboratory Earth System Model, version 4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 x 1\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGISS-E2-2-G\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGoddard Institute for Space Studies Model E2.2G\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2. x 2.5\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er2i1p1f2\u003c/p\u003e \u003cp\u003er1i1p1f2*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMPI-ESM1-2-LR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMax Planck Institute Earth System Model, version 1.2, low resolution\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.9 x 1.9\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er9i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMPI-ESM1-2-HR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMax Planck Institute Earth System Model, version 1.2, high resolution\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.9 x 0.9\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2 x WRF\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMRI-ESM2-0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMeteorological Research Institute Earth System Model, version 2.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.125 x 1.125\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eNorESM2-MM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eNorwegian Earth System Model, version 2, 1 degree resolution\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1 x 1\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBARPA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2 x WRF\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCAM_oc\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUKESM1-0-LL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnited Kingdom Earth System Model, version 1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.875 x 1.25\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003er1i1p1f2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2 x WRF\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eQldFCP-2 and CCAM-ACS used the stretched grid Conformal Cubic Atmospheric Model CCAM; Thatcher, 2020) for downscaling 11 and 7 GCMs, respectively. QldFCP-2 downscaled to a 10 km spatial resolution over Australia, and used bias and variance corrected sea surface temperatures (SSTs) and sea ice, following the approach outlined by Hoffman et al.\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. Five of the QldFCP-2 CCAM simulations were run using dynamic atmosphere-ocean coupling, while the rest were run in atmosphere-only mode (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). CCAM-ACS downscaled to a 12.5 km spatial resolution and employed spectral nudging to constrain the model to follow the host GCM\u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e, with all models run using dynamic atmosphere-ocean coupling. Both QldFPC-2 and CCAM-ACS used similar configurations for atmosphere, ocean, land-surface and aerosol parametrizations, and so the differences between them are mainly due to downscaling design (spectral nudging vs bias-corrected SSTs and sea ice), different GCMs, and differences in resolution. The Bureau of Meteorology Atmospheric Regional Projections for Australia (BARPA) is an RCM based on the UK Met Office Unified Model and Joint UK Land Environment Simulator (JULES) but configured for Australia and has been applied to downscale 7 GCMs to a 15 km spatial resolution\u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e. Similar to CCAM-ACS, nudging was used to constrain the model to follow the host GCM\u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e. NARCliM2.0 (New South Wales and Australian Regional Climate Modelling) employed two configurations of the Weather Research and Forecasting (WRF) model, each adopting different parameterisations of physics to downscale 5 GCMs. The two different sets of RCM parameterizations were selected based on their ability to simulate Australia\u0026rsquo;s recent climate and statistical independence from a larger set of 78 structurally different configurations. NARCliM2.0 provides simulations at a 20 km spatial resolution over Australasia. BARPA and NARCliM2.0 follow a limited area modelling approach, and as such these RCMs were forced at the lateral boundaries and used sea surface temperature from the host models. In total, a large ensemble consisting of 39 different regional models were considered in this study across the 4 modelling groups. Further details on the individual downscaling experiment designs can be found in\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan additionalcitationids=\"CR35\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDaily observed gridded precipitation data with a spatial resolution of 0.05\u0026deg; (approximately 5 km) were obtained from the Australian Gridded Climate Data Project (AGCD\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e). Prior to analysis all datasets, including GCMs, RCMs, and observations were re-gridded to the same spatial resolution (i.e. 10 km) using distance weighting interpolation. We assessed the performance of host and downscaled models against the observational data using daily mean, 99th percentile precipitation (p99), and 99.7th percentile precipitation (p99.7). Regions with poor observational data quality during the comparison period were masked out of the model evaluation (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Model performance using these metrics was evaluated over Australia and over the eight greater capital city regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e) using the Mean Absolute Percentage Error (MAPE) and the Root Mean Square Error (RMSE) to quantify differences.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Extreme value analysis\u003c/h2\u003e \u003cp\u003eExtreme value analysis was applied to assess changes to the probability distribution of rare extreme events. Here these events represent the 1 in 10, 1 in 50, and 1 in 100 Annual Exceedance Probability (AEP) which approximately correspond to events with annual return intervals of 10 years, 50 years, and 100 years respectively. We sampled the daily timeseries of precipitation at each grid cell using the block maxima approach to derive annual maxima (AM) precipitation and then pooled together data from nearby cells using a 5x5 box centred on each grid cell to extend the data series used for the extreme event analysis\u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e. For the analysis of capital cities, we chose to pool together all the AM data within each of the regions for the analysis. The Generalised Extreme Value (GEV) distribution was then fitted to the AM series using the L-moments method for parameter estimation. The GEV distribution is a generalised expression combining the Gumbel, Fr\u0026eacute;chet, and the Weibull distributions and is given by:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:G\\left(x\\right)=exp\\left\\{-{\\left[1+\\xi\\:\\left(\\frac{x-\\mu\\:}{\\sigma\\:}\\right)\\right]}^{\\frac{-1}{\\xi\\:}}\\right\\},\\:for\\:\\left\\{x:1+\\xi\\:\\left(\\frac{x-\\mu\\:}{\\sigma\\:}\\right)\u0026gt;0\\right\\}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere, \u003cem\u003e\u0026micro;\u003c/em\u003e, \u003cem\u003eσ\u003c/em\u003e, and \u003cem\u003eξ\u003c/em\u003e are the location, scale, and shape parameters, respectively. Here, the location parameter is a measure of the central tendency and is loosely linked to the mean, the scale parameter is a measure of variance, and the shape parameter describes the tail behaviour.\u003c/p\u003e \u003cp\u003eThe shape parameter is particularly important for the estimation of rare extremes, as it describes the behaviour of the tail. When \u003cem\u003eξ\u003c/em\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0 the distribution is unbounded with no upper limit, while when \u003cem\u003eξ\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0 the distribution is bounded by an upper limit\u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. Accurate estimation of the shape parameter necessitates many years of data to adequately fit, as it is susceptible to outliers\u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e. We fit the GEV distribution to two 30-year periods representing the recent past (1981\u0026ndash;2010) which we term the reference period, and the far future (2071\u0026ndash;2100). However, as the data are pooled from nearby cells, we effectively increase the number of data points used in the analysis from 30 per grid cell to 750\u003csup\u003e71\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Climate change assessment\u003c/h2\u003e \u003cp\u003eWe examined the impacts of climate change on precipitation extremes by the end of the century (2071\u0026ndash;2100) relative to the 1981\u0026ndash;2010 reference period. We calculated the precipitation for the 1 in 10, 50, and 100 AEPs. We also evaluated changes to mean and moderately extreme (p99 and p99.7) precipitation to determine changes across a range of precipitation intensities. The 99.7th percentile precipitation was evaluated as this approximately corresponds to an event which would occur once per year. Calculations were applied at each individual grid cell for each of the climate models considered. The results for each of the projections were assessed individually and by model ensemble (i.e., 4 groups of RCMs and 1 group of GCMs), including a combined ensemble of all downscaled projections (CORDEX ensemble) assessed for the greater capital city regions. Model ensemble averages and medians were calculated, with calculations of the averages adopting a one model one vote rule. This weights the models according to the number of simulations per host model (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) and results in an 11-model average for QldFCP-2, a 5-model average for NARCliM2.0, and a 7-model model average for CCAM-ACS and BARPA-ACS. For the CMIP6 GCMs we also weighted the two EC-Earth3 and MPI-ESM1-2 models to give a 14-model weighted ensemble. Resulting changes are presented as spatial maps over Australia using the ensemble average or median change and as boxplots for the eight greater capital city regions assessed using all model projections to better understand uncertainty. Averages were used for the analysis of mean, p99, and p99.7, while medians were used for the results of the GEV extremes to ensure they were not influenced by outliers.\u003c/p\u003e \u003cp\u003eWe present precipitation changes as a rate per degree of global temperature change (i.e., the global mean temperature, including both land and ocean regions). For consistency, we used the global temperature changes derived from the host GCMs as opposed to warming from the downscaled projections. This was implemented to ensure consistency between the different model ensembles and due to difficulty in deriving global warming levels from limited area RCMs such as BARPA and WRF. We calculated the projected changes per degree warming for the high emissions scenario (SSP370) only, as only this scenario and SSP126 were shared by all modelling groups, and as SSP370 would show a more sensitive response by the end of the century. Scaling rates were calculated for each model from the projected changes between the reference period (1981\u0026ndash;2010) and the end of the century (2071\u0026ndash;2100) as per the equation below.\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\frac{\\%\\:change}{degree\\:warming}=\\:\\left[\\left\\{{\\left(\\frac{{P}_{f}}{{P}_{r}}\\right)}^{\\left(\\frac{1}{{T}_{f}-{T}_{r}}\\right)}\\right\\}-1\\right]\\times\\:100$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003ef\u003c/em\u003e\u003c/sub\u003e is the precipitation in the future period, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003er\u003c/em\u003e\u003c/sub\u003e is the precipitation in the reference period, \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ef\u003c/em\u003e\u003c/sub\u003e is the mean global temperature in the future period, and \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003er\u003c/em\u003e\u003c/sub\u003e is the mean global temperature in the reference period. To determine where there is confidence in the scaling rate change over Australia, we adopt the signal-to-noise ratio to see where the climate change signal emerges over the \u0026lsquo;noise\u0026rsquo; for each of the climate model ensembles considered\u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e. Here, we consider the noise as the standard deviation from all the models in each ensemble. Stippling is shown on the ensemble mean and median change maps where the signal-to-noise ratio is greater than 1.0, as this is a commonly adopted threshold used in the literature\u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e,\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e. To compare differences between the downscaled models and host models, we also compared the Probability Density Function (PDF) plot of the changes from these RCMs and GCMs. Here, the downscaled projections were also weighted evenly between modelling groups to avoid biasing the results from the larger downscaled ensembles.\u003c/p\u003e \u003c/div\u003e\n"},{"header":"Declarations","content":"\u003ch2\u003eData availability\u003c/h2\u003e\n\u003cp\u003eAll data used in this study are publicly available. The downscaled climate projections which contribute to the CORDEX-CMIP6 Australasia domain can be accessed through the National Computer Infrastructure: https://nci.org.au/. The QldFCP-2 dataset is available at: https://dx.doi.org/10.25914/8fve-1910. The CCAM-ACS dataset is available at: https://dx.doi.org/10.25914/3r9s-pb86. The BARPA dataset is available at: https://dx.doi.org/10.25914/z1x6-dq28. The NARCliM2.0 dataset is available at: https://dx.doi.org/10.25914/3r9s-pb86. The CMIP6 global climate model data are available through the Earth System Grid Federation at: http://esgf.llnl.gov/. Gridded AGCD observations can be accessed at: https://dx.doi.org/10.25914/6009600b58196.\u003c/p\u003e\n\u003ch2\u003eCode availability\u003c/h2\u003e\n\u003cp\u003eAll relevant codes used in this work are available upon request from the corresponding author.\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eAll authors declare no financial or non-financial competing interests.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eR.E.: Writing \u0026ndash; Original draft preparation, Conceptualization, Methodology, Formal analysis. J.S.: Conceptualization, Data Curation, Methodology. R.T.: Conceptualization, Methodology, Writing - Review \u0026amp; Editing. S.C.: Data Curation, Writing - Review \u0026amp; Editing. C.W.: Conceptualization, Methodology, Writing - Review \u0026amp; Editing. J.E.: Conceptualization, Methodology, Writing - Review \u0026amp; Editing. M.T.: Data Curation, Writing - Review \u0026amp; Editing. G.D.V.: Data Curation, Writing - Review \u0026amp; Editing. C.S.: Data Curation, Writing - Review \u0026amp; Editing\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eWe acknowledge support by Lindsay Brebber from Information and Digital Science Delivery of the Department of Environment and Science for support with high performance computing and data storage.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAlexander, L. V. \u0026amp; Arblaster, J. M. Historical and projected trends in temperature and precipitation extremes in Australia in observations and CMIP5. Weather Clim. Extrem. 15, 34\u0026ndash;56 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eContractor, S., Donat, M. G. \u0026amp; Alexander, L. V. Intensification of the Daily Wet Day Rainfall Distribution Across Australia. Geophys. Res. Lett. 45, 8568\u0026ndash;8576 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGroisman, P. Y. \u003cem\u003eet al.\u003c/em\u003e Trends in intense precipitation in the climate record. J. Clim. 18, 1326\u0026ndash;1350 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJayaweera, L., Wasko, C., Nathan, R. \u0026amp; Johnson, F. Non-stationarity in extreme rainfalls across Australia. J. Hydrol. 624, 129872 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEccles, R., Zhang, H. \u0026amp; Hamilton, D. A review of the effects of climate change on riverine flooding in subtropical and tropical regions. J. Water Clim. Change 10, 687\u0026ndash;707 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, X. \u003cem\u003eet al.\u003c/em\u003e Global Projection of Flood Risk With a Bivariate Framework Under 1.5\u0026ndash;3.0\u0026deg;C Warming Levels. Earths Future 12, e2023EF004312 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWestra, S., Alexander, L. V. \u0026amp; Zwiers, F. W. Global Increasing Trends in Annual Maximum Daily Precipitation. (2013) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/JCLI-D-12-00502.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-12-00502.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMyhre, G. \u003cem\u003eet al.\u003c/em\u003e Frequency of extreme precipitation increases extensively with event rareness under global warming. Sci. Rep. 9, 16063 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrenberth, K. E., Dai, A., Rasmussen, R. M. \u0026amp; Parsons, D. B. The Changing Character of Precipitation. Bull. Am. Meteorol. Soc. 84, (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePfahl, S., O\u0026rsquo;Gorman, P. A. \u0026amp; Fischer, E. M. Understanding the regional pattern of projected future changes in extreme precipitation. Nat. Clim. Change 7, 423\u0026ndash;427 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFowler, H. J. \u003cem\u003eet al.\u003c/em\u003e Anthropogenic intensification of short-duration rainfall extremes. Nat. Rev. Earth Environ. 2, 107\u0026ndash;122 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO\u0026rsquo;Gorman, P. A. Precipitation Extremes Under Climate Change. Curr. Clim. Change Rep. 1, 49\u0026ndash;59 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePendergrass, A. G. What precipitation is extreme? Science 360, 1072\u0026ndash;1073 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrose, M. R. \u003cem\u003eet al.\u003c/em\u003e Insights From CMIP6 for Australia\u0026rsquo;s Future Climate. Earths Future 8, e2019EF001469 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGr\u0026uuml;ndemann, G. J., van de Giesen, N., Brunner, L. \u0026amp; van der Ent, R. Rarest rainfall events will see the greatest relative increase in magnitude under future climate change. Commun. Earth Environ. 3, 1\u0026ndash;9 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, C. \u003cem\u003eet al.\u003c/em\u003e Changes in Annual Extremes of Daily Temperature and Precipitation in CMIP6 Models. (2021) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/JCLI-D-19-1013.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-19-1013.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbdelmoaty, H. M. \u0026amp; Papalexiou, S. M. Changes of Extreme Precipitation in CMIP6 Projections: Should We Use Stationary or Nonstationary Models? J. Clim. 36, 2999\u0026ndash;3014 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReder, A., Raffa, M., Montesarchio, M. \u0026amp; Mercogliano, P. Performance evaluation of regional climate model simulations at different spatial and temporal scales over the complex orography area of the Alpine region. Nat. Hazards 102, 151\u0026ndash;177 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBo\u0026eacute;, J. \u0026amp; Terray, L. Land\u0026ndash;sea contrast, soil-atmosphere and cloud-temperature interactions: interplays and roles in future summer European climate change. Clim. Dyn. 42, 683\u0026ndash;699 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChapman, S. \u003cem\u003eet al.\u003c/em\u003e Evaluation of Dynamically Downscaled CMIP6-CCAM Models Over Australia. Earths Future 11, e2023EF003548 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrose, M. R. \u003cem\u003eet al.\u003c/em\u003e The role of topography on projected rainfall change in mid-latitude mountain regions. Clim. Dyn. 53, 3675\u0026ndash;3690 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian, T. \u003cem\u003eet al.\u003c/em\u003e Resolved complex coastlines and land\u0026ndash;sea contrasts in a high-resolution regional climate model: a comparative study using prescribed and modelled SSTs. Tellus Dyn. Meteorol. Oceanogr. 65, 19951 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiorgi, F. Thirty years of regional climate modeling: Where are we and where are we going next? J. Geophys. Res. Atmospheres 124, 2018JD030094 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGutowski, W. J. \u003cem\u003eet al.\u003c/em\u003e The Ongoing Need for High-Resolution Regional Climate Models: Process Understanding and Stakeholder Information. Bull. Am. Meteorol. Soc. 101, E664\u0026ndash;E683 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, X. \u003cem\u003eet al.\u003c/em\u003e Indices for monitoring changes in extremes based on daily temperature and precipitation data. Wiley Interdiscip. Rev. Clim. Change 2, 851\u0026ndash;870 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBao, J., Sherwood, S. C., Alexander, L. V. \u0026amp; Evans, J. P. Future increases in extreme precipitation exceed observed scaling rates. Nat. Clim. Change 7, 128\u0026ndash;132 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChapman, S., Syktus, J., Trancoso, R., Toombs, N. \u0026amp; Eccles, R. Projected changes in mean climate and extremes from downscaled high-resolution CMIP6 simulations in Australia. Weather Clim. Extrem. 46, 100733 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJi, F. \u003cem\u003eet al.\u003c/em\u003e Introducing NARCliM1.5: Evaluation and projection of climate extremes for southeast Australia. Weather Clim. Extrem. 38, 100526 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerold, N. \u003cem\u003eet al.\u003c/em\u003e Projected changes in the frequency of climate extremes over southeast Australia. Environ. Res. Commun. 3, 011001 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMantegna, G. A., White, C. J., Remenyi, T. A., Corney, S. P. \u0026amp; Fox-Hughes, P. Simulating sub-daily Intensity-Frequency-Duration curves in Australia using a dynamical high-resolution regional climate model. J. Hydrol. 554, 277\u0026ndash;291 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWasko, C., Guo, D., Ho, M., Nathan, R. \u0026amp; Vogel, E. Diverging projections for flood and rainfall frequency curves. J. Hydrol. 620, 129403 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrancoso, R. \u003cem\u003eet al.\u003c/em\u003e Significantly wetter or drier future conditions for one to two thirds of the world\u0026rsquo;s population. Nat. Commun. 15, 483 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKendon, E. J., Rowell, D. P., Jones, R. G. \u0026amp; Buonomo, E. Robustness of future changes in local precipitation extremes. J. Clim. 21, 4280\u0026ndash;4297 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Virgilio, G. \u003cem\u003eet al.\u003c/em\u003e Design, evaluation, and future projections of the NARCliM2.0 CORDEX-CMIP6 Australasia regional climate ensemble. Geosci. Model Dev. 18, 671\u0026ndash;702 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward, E. \u003cem\u003eet al.\u003c/em\u003e Evaluation of multi-season convection-permitting atmosphere \u0026ndash; mixed-layer ocean simulations of the Maritime Continent. Geosci. Model Dev. 17, 3815\u0026ndash;3837 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchroeter, B. J. E., Ng, B., Takbash, A., Rafter, T. \u0026amp; Thatcher, M. A Comprehensive Evaluation of Mean and Extreme Climate for the Conformal Cubic Atmospheric Model (CCAM). (2024) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/JAMC-D-24-0004.1\u003c/span\u003e\u003cspan address=\"10.1175/JAMC-D-24-0004.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLavender, S. L. \u0026amp; Abbs, D. J. Trends in Australian rainfall: contribution of tropical cyclones and closed lows. Clim. Dyn. 40, 317\u0026ndash;326 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWasko, C. \u003cem\u003eet al.\u003c/em\u003e A systematic review of climate change science relevant to Australian design flood estimation. Hydrol. Earth Syst. Sci. Discuss. 1\u0026ndash;48 (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.5194/hess-2023-232\u003c/span\u003e\u003cspan address=\"10.5194/hess-2023-232\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWasko, C., Lu, W. T. \u0026amp; Mehrotra, R. Relationship of extreme precipitation, dry-bulb temperature, and dew point temperature across Australia. Environ. Res. Lett. 13, 074031 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShiogama, H. \u003cem\u003eet al.\u003c/em\u003e Important distinctiveness of SSP3\u0026ndash;7.0 for use in impact assessments. Nat. Clim. Change 13, 1276\u0026ndash;1278 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, L., Wang, Z., Xu, Y. \u0026amp; Fu, Q. Sensitivity of precipitation extremes to radiative forcing of greenhouse gases and aerosols. Geophys. Res. Lett. 43, 9860\u0026ndash;9868 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, L., Wang, Z., Xu, Y., Fu, Q. \u0026amp; Dong, W. Larger Sensitivity of Precipitation Extremes to Aerosol Than Greenhouse Gas Forcing in CMIP5 Models. J. Geophys. Res. Atmospheres 123, 8062\u0026ndash;8073 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBell, S. S. \u003cem\u003eet al.\u003c/em\u003e Projections of southern hemisphere tropical cyclone track density using CMIP5 models. Clim. Dyn. 52, 6065\u0026ndash;6079 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLavender, S. L. \u0026amp; Walsh, K. J. E. Dynamically downscaled simulations of Australian region tropical cyclones in current and future climates. Geophys. Res. Lett. 38, 1\u0026ndash;6 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalsh, K. \u003cem\u003eet al.\u003c/em\u003e Natural hazards in Australia: storms, wind and hail. Clim. Change 139, 55\u0026ndash;67 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDowdy, A. J. \u003cem\u003eet al.\u003c/em\u003e Review of Australian east coast low pressure systems and associated extremes. Clim. Dyn. 53, 4887\u0026ndash;4910 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePepler, A. S. \u0026amp; Dowdy, A. J. Australia\u0026rsquo;s Future Extratropical Cyclones. J. Clim. 35, 7795\u0026ndash;7810 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePepler, A. S. \u003cem\u003eet al.\u003c/em\u003e Projected changes in east Australian midlatitude cyclones during the 21st century. Geophys. Res. Lett. 43, 334\u0026ndash;340 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDowdy, A. J. Climatology of thunderstorms, convective rainfall and dry lightning environments in Australia. Clim. Dyn. 54, 3041\u0026ndash;3052 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrein, A. F. \u003cem\u003eet al.\u003c/em\u003e The future intensification of hourly precipitation extremes. Nat. Clim. Change 7, 48\u0026ndash;52 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu, C.-H. \u003cem\u003eet al. BARRA-C2: Development of the Kilometre-Scale Downscaled Atmospheric Reanalysis over Australia\u003c/em\u003e. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFosser, G. \u003cem\u003eet al.\u003c/em\u003e Convection-permitting climate models offer more certain extreme rainfall projections. Npj Clim. Atmospheric Sci. 7, 1\u0026ndash;10 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeloitte. \u003cem\u003eBuilding Resilience to Natural Disasters in Our States and Territories\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://australianbusinessroundtable.com.au/assets/documents/ABR_building-resilience-in-our-states-and-territories.pdf\u003c/span\u003e\u003cspan address=\"https://australianbusinessroundtable.com.au/assets/documents/ABR_building-resilience-in-our-states-and-territories.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWasko, C., Sharma, A. \u0026amp; Pui, A. Linking temperature to catastrophe damages from hydrologic and meteorological extremes. J. Hydrol. 602, 126731 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWright, D. B., Bosma, C. D. \u0026amp; Lopez-Cantu, T. U.S. Hydrologic Design Standards Insufficient Due to Large Increases in Frequency of Rainfall Extremes. Geophys. Res. Lett. 46, 8144\u0026ndash;8153 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHettiarachchi, S., Wasko, C. \u0026amp; Sharma, A. Increase in flood risk resulting from climate change in a developed urban watershed \u0026ndash; the role of storm temporal patterns. Hydrol. Earth Syst. Sci. 22, 2041\u0026ndash;2056 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrader, S. M. \u0026amp; Ashley, W. S. The Expanding Bull\u0026rsquo;s-Eye Effect. Weatherwise 68, 23\u0026ndash;29 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWing, O. E. J. \u003cem\u003eet al.\u003c/em\u003e Estimates of present and future flood risk in the conterminous United States. Environ. Res. Lett. 13, 034023 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndreadis, K. M. \u003cem\u003eet al.\u003c/em\u003e Urbanizing the floodplain: global changes of imperviousness in flood-prone areas. Environ. Res. Lett. 17, 104024 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan den Honert, R. C. \u0026amp; McAneney, J. The 2011 Brisbane floods: causes, impacts and implications. Water 3, 1149\u0026ndash;1173 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDykman, C., Sharma, A., Wasko, C. \u0026amp; Nathan, R. Can annual streamflow volumes be characterised by flood events alone? J. Hydrol. 617, 128884 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEccles, R., Zhang, H., Hamilton, D., Trancoso, R. \u0026amp; Syktus, J. Impacts of climate change on nutrient and sediment loads from a subtropical catchment. J. Environ. Manage. 345, 118738 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Virgilio, G. \u003cem\u003eet al.\u003c/em\u003e Selecting CMIP6 GCMs for CORDEX Dynamical Downscaling: Model Performance, Independence, and Climate Change Signals. Earths Future 10, 1\u0026ndash;24 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrose, M. R. \u003cem\u003eet al.\u003c/em\u003e A CMIP6-based multi-model downscaling ensemble to underpin climate change services in Australia. Clim. Serv. 30, 100368 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrancoso, R., Syktus, J., Toombs, N. \u0026amp; Chapman, S. \u003cem\u003eAssessing and Selecting CMIP6 GCMs Ensemble Runs Based on Their Ability to Represent Historical Climate and Future Climate Change Signal\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://meetingorganizer.copernicus.org/EGU23/EGU23-11412.html\u003c/span\u003e\u003cspan address=\"https://meetingorganizer.copernicus.org/EGU23/EGU23-11412.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.5194/egusphere-egu23-11412\u003c/span\u003e\u003cspan address=\"10.5194/egusphere-egu23-11412\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoffmann, P., Katzfey, J. J., McGregor, J. L. \u0026amp; Thatcher, M. Bias and variance correction of sea surface temperatures used for dynamical downscaling. J. Geophys. Res. 121, 12,877\u0026thinsp;\u0026ndash;\u0026thinsp;12,890 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThatcher, M. \u0026amp; McGregor, J. L. Using a Scale-Selective Filter for Dynamical Downscaling with the Conformal Cubic Atmospheric Model. Mon. Weather Rev. 137, 1742\u0026ndash;1752 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu, C.-H. \u003cem\u003eet al. BARPA: New Development of ACCESS-Based Regional Climate Modelling for Australian Climate Service\u003c/em\u003e. (Bureau of Meteorology, 2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStassen, C. \u003cem\u003eet al. Development and Assessment of Regional Atmospheric Nudging in ACCESS\u003c/em\u003e. (Bureau of Meteorology Melbourne, 2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEvans, A., Jones, D., Smalley, R. \u0026amp; Lellyett, S. An enhanced gridded rainfall analysis scheme for Australia. Aust. Bur. Meteorol. Melb. VIC Aust. 66, 55\u0026ndash;67 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, J., Evans, J., Johnson, F. \u0026amp; Sharma, A. A comparison of methods for estimating climate change impact on design rainfall using a high-resolution RCM. J. Hydrol. 547, 413\u0026ndash;427 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColes, S. \u003cem\u003eAn Introduction to Statistical Modeling of Extreme Values\u003c/em\u003e. (Springer-Verlag, Bristol, UK, 2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePapalexiou, S. M. \u0026amp; Koutsoyiannis, D. Battle of extreme value distributions: A global survey on extreme daily rainfall. Water Resour. Res. 49, 187\u0026ndash;201 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHawkins, E. \u0026amp; Sutton, R. The potential to narrow uncertainty in projections of regional precipitation change. Clim. Dyn. 37, 407\u0026ndash;418 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChapman, S., Syktus, J., Trancoso, R., Toombs, N. \u0026amp; Eccles, R. Projected Changes in Mean Climate and Extremes from Downscaled High-Resolution Cmip6 Simulations in Australia. SSRN Scholarly Paper at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2139/ssrn.4836517\u003c/span\u003e\u003cspan address=\"10.2139/ssrn.4836517\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-natural-hazards","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [npj Natural Hazards](https://www.nature.com/npjnathazards/)","snPcode":"44304","submissionUrl":"https://submission.springernature.com/new-submission/44304/3","title":"npj Natural Hazards","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Downscaled climate projections, Extreme precipitation uncertainty, Generalised Extreme Value distribution, Global Warming Levels, Precipitation extremes, Regional climate modelling","lastPublishedDoi":"10.21203/rs.3.rs-6291947/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6291947/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eExtreme precipitation events are widely held to become more intense and frequent as a result of climate change, which will have major impacts for future flooding with implications for the environment, infrastructure, agriculture, and human life. We investigated projected changes to daily mean, moderately extreme (99th and 99.7th percentile), and rare extreme (Annual Exceedance Probability (AEP) 1 in 10, 50, and 100) precipitation events across Australia and its greater capital cities, where approximately two thirds of the Australian population reside. We used dynamically downscaled CMIP6 precipitation simulations from 4 modelling groups in Australia. This large ensemble consists of 19 different host models downscaled using 3 distinct regional climate models in 5 different configurations, making an ensemble of 39 different downscaled simulations. The changes in mean and extreme precipitation events were quantified at each grid cell from each of the models according to the rate of change per degree of global warming. The largest increases to precipitation extremes were seen over northern Australia, with the 1 in 100 AEP event in Darwin projected to increase by 11.9% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 12.2% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the downscaled and host ensemble averages, respectively. Other capital cities had lower increases but still substantial (7.6% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for Brisbane, 7.3% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for Sydney, 3.4% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for Melbourne, and 4.4% K\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for Perth). Large spatial differences were noted among the downscaled ensembles, with models from different modelling groups showing varying spatial patterns and magnitudes of change. These results highlight the influence of the downscaling approach in determining changes to precipitation extremes and show the need to consider large ensembles to ensure uncertainties in host models and downscaling methods can be accounted for. The findings can inform decision making around flood management, urban planning, urban water supply and agriculture around Australia, in addition to revealing globally relevant scientific insights.\u003c/p\u003e","manuscriptTitle":"Substantial increases in future precipitation extremes – insights from a large ensemble of downscaled CMIP6 models","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-21 10:41:06","doi":"10.21203/rs.3.rs-6291947/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-04-19T04:17:27+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-14T17:15:18+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-10T15:01:01+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"183328911630067212775230893610955736091","date":"2025-04-03T13:27:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"211984406985782643422919591989961274717","date":"2025-04-01T18:04:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"42253361386825006490474743758253958551","date":"2025-04-01T16:25:42+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-31T20:16:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-03-24T22:03:12+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-24T19:53:57+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Natural Hazards","date":"2025-03-24T05:38:18+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-natural-hazards","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [npj Natural Hazards](https://www.nature.com/npjnathazards/)","snPcode":"44304","submissionUrl":"https://submission.springernature.com/new-submission/44304/3","title":"npj Natural Hazards","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1eb6db9d-36e7-41ba-9b61-a3193ea873f5","owner":[],"postedDate":"April 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":47383565,"name":"Earth and environmental sciences/Climate sciences/Climate change/Projection and prediction"},{"id":47383566,"name":"Earth and environmental sciences/Climate sciences/Hydrology"},{"id":47383567,"name":"Earth and environmental sciences/Natural hazards"}],"tags":[],"updatedAt":"2025-06-13T06:38:37+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-21 10:41:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6291947","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6291947","identity":"rs-6291947","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}