Using historical tropical cyclone climate datasets to examine wind speed recurrence for coastal Australia

Using historical tropical cyclone climate datasets to examine wind speed recurrence for coastal Australia | 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 Using historical tropical cyclone climate datasets to examine wind speed recurrence for coastal Australia S. S. Bell, A. J. Dowdy, H. A. Ramsay, S. S. Chand, C-H Su, H. Ye This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-1507699/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract Likelihood estimates of extreme winds, including those from tropical cyclones (TCs) at certain locations are used to inform wind load standards for structural design. Here, wind speed average recurrence intervals (ARIs) determined from TC climate data dating back to the 1970s in two quantile-quantile adjusted reanalysis datasets (ERA5 and BARRA [1990]), and best-track observations for context, were compared with Standardized ARIs (AS/NZS) across seven tropical and two subtropical Australian inland coastal regions. Inherent differences between the data used to determine the Standard ARIs (large sample size allow for larger extrapolations; GEV function) and TC data ARIs (smaller sample size and less certain data; the more asymptotic Lognormal/Weibull functions are used) led to the use of different extreme value functions. Results indicated that although these are two distinct ways of determining design wind speeds, when they are considered equivalent, there was a moderate reproduction of the ARI curves with respect to the Standard in both reanalysis datasets, suggesting that similar analyses using climate model products can provide useful information on these types of metrics with some caveats. Trends in TC wind strength affecting coastal Australia were also analyzed, indicating a potential slight downtrend in tropical West coast TC wind strength and slight uptrends for tropical East coast TC wind strength, subtropical West coast TC wind strength and subtropical East coast ETC wind strength with low-to-medium confidence. Noting considerable uncertainties, and large interannual variability makes it hard to determine any of these results with a high degree of confidence based on current data. It is also difficult to attribute any noted changes in extreme wind gust strength in these datasets to anthropogenic climate change at this time given the relatively short time period and limitations of the data quality including that over longer time periods and preceding the era of widespread satellite coverage in the Australian region. It is hoped that climate models can provide a more homogenous base for these types of evaluations and subsequent projections with respect to climate change simulations. tropical cyclones extratropical cyclones wind speed return periods coastal hazard reanalysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction Extreme wind speeds induced by tropical cyclones (TCs) can have considerable impacts on coastal communities and infrastructure situated in tropical Australia. In light of anthropogenic-induced climate change, it is important to investigate the potential evolution of TC-wind hazards for the benefit of future planning and disaster risk management in Australia. The expected average recurrence interval (ARI) of extreme wind speeds is a metric widely used by engineers and climate scientists alike to gauge the wind hazard for specific regions. In Australia, this is described by the “structural design actions” for winds 1 herein referred to as “the Standard”. Different regions of Australia are assigned various designations (A, B, C or D, in ascending order of strength) based on the historical wind strength (ARI) observed in that region. For example, Australia’s tropical and subtropical coastlines have much higher designations (B–D) than the southern and inland regions (A) due to the increased exposure to TCs. Historically, appropriate regional wind speed designations have been determined by weather station surface anemometer measurements (corrected for terrain, topography and gust duration) and to a lesser extent damage surveys from past events (including in some cases back calculations of maximum wind speeds from failed simple structures such as road signs) and these methods remain very useful. In the latest edition of the AS/NZS structural design Standard (1170.2:2021), for the first time, a climate change multiplier (1.05) was included as mandatory to consider in some regions of Australia. This is consistent with a broad review and synthesis of available lines of evidence that indicate an increase in TC intensity with projected future warming is more likely than a decrease 2,3 . For local context, the East coast of Australia has experienced a recent influx of severe Australian Category 4/5 TCs 4 (Table 1 ) while the West coast of Australia has experienced fewer Category 4/5 TCs recently 5 . This leads to the question of what role can climate model experiments play in informing us of likely future changes in fine-scale metrics such as wind speed ARIs that have traditionally come from observed ground truth data. Climate models, despite having their own limitations such as biases and inaccurate representations of some small- and large-scale physical processes, still provide homogenous simulations that can be useful for examining the potential influence of climate change now and into the future, as a complementary line of evidence to observations-based data (which also have limitations including due to evolving operational practices for wind observations in Australia with a relatively short time period of high-quality data being available). Table 1 The Australian TC wind scale as defined by wind gust speed (3-second). Category Gust speed 1 24.5–34 m s − 1 2 34–45.5 m s − 1 3 45.5–61.6 m s − 1 4 61.6–77.4 m s − 1 5 > 77.4 m s − 1 It is timely that with the recent release of new, finer resolution reanalysis products, including the Bureau of Meteorology Atmospheric high-resolution Regional Reanalysis for Australia (BARRA 6 ) and the global ERA5 reanalysis 7 produced by the ECMWF; the feasibility of determining wind speed ARIs in climate model type products can be tested. This can be achieved by direct comparison with the ground truth determined wind speed ARIs from AS/NZS 1170.2. The two reanalysis datasets can also provide additional lines of evidence in relation to current historical TC trends and extreme wind speeds affecting Australia. The simulation of peak wind gust speeds from TCs in reanalysis products poses many challenges, and the products used here do not completely resolve the absolute wind gust speeds of the more intense TCs 2,8 . However, by calibrating the reanalysis data to the observations, such as from Automatic Weather Stations (AWS) and from the best track estimates of TC wind speeds 9,10,11 useful information on absolute gust speeds in reanalysis data can be obtained, though with some obvious limitations 12 . That is, we are not suggesting that the ARIs calculated here should replace the current structural AS/NZS wind speed Standard, but rather to serve as an additional line of evidence contributing to a more comprehensive understanding of potential long-term changes in this metric and in the risk of TC-wind hazards for coastal Australia. In particular, several data sets are used here for the first time to examine TCs and associated extreme wind speeds in these regions, thereby providing new insight on this important topic, which is intended to help provide useful guidance complementary to existing information. Here, we evaluate ARIs of TC-related daily-maximum wind gusts over eight Australian coastal regions (Fig. 1 ). These regions closely correspond with those designated by AS/NZS 1170.2:2021, allowing for a comparison of these wind speeds to the Standard. Two subtropical regions on either side of the Australian continent, where weakening TCs and extratropical cyclones (ETCs) can sometimes occur, are additionally examined. Examination of these two regions may provide some guidance on a key scientific question around the potential poleward movement of TCs and their wind hazards. The following section provides details on the Data and Methods used. Results are shown in Section 3, and a Summary and Conclusion is provided in Section 4. 2. Data And Methods 2.1. Data 2.1.1 Best track The Bureau of Meteorology (BoM) TC dataset (available online at http://www.bom.gov.au/cyclone/tropical-cyclone-knowledge-centre/databases/ ) includes historical TCs that have occurred in the Australian region (from 90 to 160 degrees East). Since the introduction of operational satellite technology in the early 1970s, the historical records of TCs (analyzed here over the period 1971 to 2021), have included best track sustained wind speed, wind gust estimates, and TC position and time information (with a degree of uncertainty 13 ). We note that some TCs in the most recent two years of the dataset are considered provisional, meaning they have not yet undergone formal best track review; and that several TCs in the early 1970s do not have wind speed information available. TC wind speed estimates in the best track data have been most widely determined by the satellite-based Dvorak technique 9,10,14 . This technique involves the subjective assignment of Dvorak numbers to satellite imagery of TCs to give an indication of the wind speed. However, especially more recently, other tools and techniques have been used by operational forecast centres and meteorological agencies to estimate TC intensities. Such tools and techniques, which include scatterometers, AWS, as well as the Advanced Dvorak technique (ADT), and the SATellite CONsensus intensity approach (SATCON), can influence the final Dvorak number estimate (and therefore the intensity) that is eventually recorded in the best track data 11 . These new tools and techniques make the recent (~ post 2003) best-track record more accurate 15 , but also embeds a degree of inhomogeneity in the dataset when compared with the earlier periods. As such, numerous studies have recommended data from more recent periods for evaluations of TC intensity estimates 16 . For example, [ 17 ] recommends data from 1980/81 onwards for TC intensity analysis, whereas others are more conservative: recommending data from 1990 onwards 18 and from 2003 onwards 19 . In our case, we have used the entire satellite data for completeness, but placed emphasis on the period from 1990 onwards. A re-evaluation of TCs in the Australian region 19,20 provides revised wind speed (V max ) estimates in some instances compared to those in the best-track (BoM) TC dataset over the period 1981–2016, by for example, using ADT to remove subjectivity from the simple Dvorak technique. In addition, there are some cases in the best track dataset where only a TC’s sustained wind speed (i.e., 10-minute average value) was recorded in the best track TC dataset (i.e., no wind gust conversion based on an open terrain estimate is provided). Here, 3-second wind gusts were obtained by multiplying the wind speed value by a gust factor (GF) of 1.4. GFs can vary due to several factors 21 , including changes in surface roughness (where complex terrain requires a higher GF than that of the smooth open ocean 22 ). Similarly, there are also various modification factors that determine the actual wind speed a specific site will experience. These factors include the directional, shielding, terrain and topographic multipliers (e.g., AS/NZS 1170.2). Due to the broad scale of the analyses undertaken here, a GF of 1.4 was used when necessary, noting a similar value was used in most best track derivations of wind gust speed (that vary depending on terrain). As discussed in the previous paragraph, there are three options to use for representing best track wind gust values. We give a preference to the Courtney and Burton 20 adjusted V max (10-min sustained) wind speeds, where available, ahead of the best-track wind speeds such that the final TC wind gust speed values assigned to TC observations are in the following order: 1. Adjusted V max ✕ 1.4 (Courtney and Burton 20 ) (Predominantly occurring from 1980–2016) 2. Best track wind gust speed estimate (Predominantly occurring everywhere else) 3. Best track V max ✕ 1.4 (occur infrequently preceding 1980) 2.1.2 ERA5 The latest global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF), ERA5 7 , is available at the horizontal resolution of 31 km. Daily maximum wind gusts (averaged over 3-seconds at 10 m height) were calculated from hourly maximum values over the available period spanning 1979 to June 2021. Note that the effects of convection are only included in the wind gust parameterization from October 2008 onwards which embeds a degree of non-homogeneity in the data as is the case in general for data based on observations. For comparison with the Standard, the ERA5 wind speeds are calibrated to observational data (mostly derived over the ocean) using quantile-quantile scaling 12 . These winds are then further converted to standard terrain conditions (i.e., over land) by a factor of 1.05 [ 12 ]. For ARI calculations, the daily-maximum wind gusts are additionally converted to a 0.2 second averaging period (1.1), with exception of the tail 1% of quantiles. This was done to avoid exacerbating the already large absolute values of the distribution tail (due to the relatively large calibration applied on this data). Combining these conversions, the calibration factors for the original ERA5 wind data range from 1.1 (for the lowest intensity gusts) to a maximum of 1.85 for raw wind gusts > 45 m s − 1 . 2.1.3 BARRA BARRA, is a 12 km horizontal grid spacing reanalysis product covering the Australian region, (65°E – 163°W, 20°N – 60°S). This reanalysis includes assimilation of central pressure observations from the International Best Track Archive for Climate Stewardship (IBTrACS 23 ). Maximum hourly surface (10 m) wind gust speeds were acquired and used to calculate daily maximum wind gust values over the available period spanning 1990 to February 2019. There are several known issues in the BARRA data 24 , including grid-sized storms and anomalous wind gust spikes that can produce exceedingly large wind gust values at single isolated grid-points, particularly over northern Australia and near coastlines. To circumvent this issue, the hourly maximum grid cell for a specific region is discarded and replaced by the second-maximum value. The hour-on-hour increase in maximum wind gust speed is also limited to 20 ms − 1 for the final calculation of daily-maximum wind gust speeds. In cases where there is no grid-point storm or extreme gust spike present, this method usually results in a negligible reduction (< 1 m s − 1 ) in the value of the maximum wind gust speed. As was done for the ERA5 data, the BARRA reanalysis data are calibrated using quantile-quantile scaling to match the observational data while preserving the original distribution of the data 12 . Slightly different to the ERA5 data, the entire BARRA distribution is converted to an averaging period of 0.2 seconds (1.1) and a larger value is used 1,12 (1.13) to convert to standard terrain conditions. As compared to the calibration of ERA5 data, the finer resolution BARRA data required smaller calibration factors (with a maximum calibration factor of 1.35 for raw wind gusts > 70 m s − 1 ), such that this approach provides a suitable means of reducing limitations of the reanalysis products while producing calibrated data that are more consistent with the observations in absolute terms. 2.2. Methods 2.2.1 Wind speed regions Eight wind speed regions (Fig. 1 ) are defined here, where the tropical regions are denoted by their AZ/NZS 1170.2 designation “-B”, “-C” or “-D” suffix. The two subtropical regions are simply denoted by the “-S” suffix. The strongest designation in each region corresponds to the regions used here uniformly with exception of the WA-S region which lies only partly under the Region-C designation (between 25–27°S). In AS/NZS 1170.2:2021, all regions with exception of EA-S have a mandatory climate change multiplier (1.05) included in their ARI calculations. However, as this study focuses on the present and historical climate, these climate change multipliers are not considered here. 2.2.2 TC-related wind gusts Measures of wind gust maxima are confined to the eight wind speed regions as shown in Fig. 1 . Daily maximum wind gusts for these regions are considered to be “TC-related” if there is a TC track position either inside the region, or its centre is within a 4 ° radius on the same day as the daily-maximum wind gust. Such “days” that are determined to be TC-related are referred to as “TC days”. The TC data is defined from the BoM TC dataset (best track) and extra-tropical transition days as flagged in the database were not removed from the analysis. 2.2.3 Detection of ETC tracks and ETC-related wind gusts For the two subtropical regions denoted with the suffix “-S” in Fig. 1 , extratropical cyclone (ETC) days are also considered, in addition to TC days (i.e., days on which ETCs occur inside or within 4 ° radius of a given region). The cyclone tracking scheme of Murray and Simmonds 25 with refinements 26 was used to detect their occurrence in ERA5 data over the full time period and applied to both ERA5 and BARRA wind gust data. This scheme has been applied previously in reanalysis data to produce ETC tracks 27 . The number of ETCs was not tuned to any existing climatological records, and so lower intensity circulations can remain in our definition of “ETC days”. Detected ETCs were differentiated from other storms, such as detected TCs, by having a genesis location poleward of 23 ° S. 2.2.4 Extreme value modelling, ARIs and peaks The method of peak-over-threshold was used to model extreme daily-maximum cyclone-related wind gust data, where the threshold was arbitrarily chosen to produce a total average of ~ 3 peaks per year. In cases where there are insufficient peaks for a set of data/region (e.g., subtropical regions and best track data) the data is resampled with replacement to achieve the desired number of peaks. This procedure is repeated several times to ensure the most extreme values are not oversampled. Confidence bounds for ARI curves are derived by removing (lower bound) and duplicating (upper bound) the most extreme values (top 5%). The lognormal and Weibull functions best modelled the tails of our wind gust distributions in comparison to other functions (such as GEV and Gumbel; not shown). It should be noted that the Standard curves are derived using GEV, while curves derived from our data are more asymptotic (based on lognormal and Weibull shapes). The parameter estimation method of maximum likelihood was preferred although others could have been used (e.g., L-moments 28 ). In cases where the lognormal function did not provide a good fit to our data (mostly for cases with tail-heavy distributions on the West coast and for best track data), the Weibull distribution was used instead. 3. Results 3.1. Time series Time series of daily-maximum wind gusts (3-seconds) related to TCs, “non-cyclones” and “ETCs” (for the two subtropical regions) are shown for three datasets: BARRA (Fig. 2 ) ERA5 (Fig. 3 ) and Best track data (Fig. 4 ). 3.1.1 Tropical cyclone days Linear trend lines are fitted to each time series of TC day maximum wind gusts (Figs. 2 – 4 ) and none are found to be significant at the 95% confidence level (i.e., p < 0.05), with exception of best track data back to 1971 in the EA-S region (Fig. 4 ). However, the limited number of samples and large differences between trendlines beginning from different start points (1990 and 1971), embeds a lot of doubt in this result. Collectively, there are also differences in sign between at least one of the three datasets in each region. For example, if we consider that the different datasets provide a confidence range of the trendline coefficients for the combined tropical regions since 1990, these can be interpretated as: TC day maximum wind gusts on the Western Australian coast are insignificantly changing at a rate of [-0.5 * (-0.01) to 0.06] m s -1 per year. TC day maximum wind gusts on the Northern Australian coast are insignificantly changing at a rate of [-0.2 * (-0.06) to 0.1] m s -1 per year. TC day maximum wind gusts on the Eastern Australian coast are insignificantly changing at a rate of [-0.08 to (0.07) 0.5 * ] m s -1 per year. Excessively large annual trends (> 0.2 m s − 1 per year) all came from the best track dataset and are very likely influenced by inhomogeneities in that data. For these cases, the second largest annual trend from either reanalysis dataset are shown beside the original trends in curly brackets. For the two subtropical regions since 1990: TC day maximum wind gusts affecting the EA-S region have a consistent downtrend between datasets but are insignificantly changing at a rate of [-0.25 to -0.1] m s -1 per year. TC day maximum wind gusts affecting the WA-S region are insignificantly changing at a rate of [-0.07 to 0.15] m s -1 per year. When considering trendlines back to the 1970s (i.e., from ERA5 and best track data; Figs. 3 – 4 ), the trendline coefficients often become smaller, with exception of the tropical West coast regions and the subtropical regions for best track data (where trendline coefficients often become larger in size). It is also practical to consider trends in the more extreme wind gusts that are most likely to cause significant property damage (i.e., Category 3 or above; >45 m s − 1 ). Trendlines for these “extreme” TC wind gusts (not shown) in combined regions since 1990 from the two reanalysis datasets can be interpreted as: TC day maximum wind gusts on the Western Australian coast are in some cases significantly changing at a rate of [-0.53 to 0.06] m s -1 per year. TC day maximum wind gusts on the Northern Australian coast are insignificantly changing at a rate of [-0.22 to 0.14] m s -1 per year. TC day maximum wind gusts on the Eastern Australian coast are insignificantly changing at a rate of [-0.25 to 0.12] m s -1 per year. 3.1.2 Extratropical cyclone days Considering the ETC day trend lines that are fitted to the two subtropical regions (BARRA and ERA5 only; Figs. 2 – 3 ) the only significant result at the 95% confidence level (i.e., p < 0.05) was that of ETC wind gusts affecting the EA-S region since 1990 in ERA5: indicating a rate of increase of 0.2 m s − 1 per year. This result was consistent in sign but was much greater in size when compared with BARRA and additionally ERA5 data back to 1979. There was no trend in ETC wind gust strength affecting the WA-S region. For completeness, since 1990: ETC day maximum wind gusts affecting EA-S are in one case significantly changing at a rate of [-0.1 to 0.2] m s -1 per year. ETC day maximum wind gusts affecting WA-S are insignificantly changing at a rate of [0 to 0] m s -1 per year. 3.2. ARIs For comparison with the Standard wind curves, observed and model data are converted to an averaging period of 0.2 second gusts. However, this is still not a fair comparison as the Standard curves are extrapolations based on hundreds of years of data at several point locations inside a region while the reanalysis and best track curves are based on fewer years of data but at every available location (i.e., every grid cell in a region). Additionally, the reanalysis data have been calibrated to represent absolute wind speeds up to Category 5. So, it is expected that the curves based on the calibrated reanalyses (Figs. 5 – 6 ) and best track data (Fig. 7 ) would have higher returns for small-to-medium ARIs and then approach the Standard when ARI is large (> 500). In an attempt to provide a fairer comparison between our ARI curves and the Standard ARI curves, we divide the Standard ARI curves by 33 (i.e., 500-year ARIs become 15-year and so on) to get equivalent curves for comparison which are additionally displayed to the regular Standard ARIs (dashed and solid black curves in Figs. 5 – 7 ). 3.2.1 Tropical regions For the regions nearest to coastlines (WA-D, NA-C and EA-C; top rows in Figs. 5 – 7 ) the equivalent Standard curve generally lies in between or within the confidence estimates of the reanalysis and best track derived curves. However, this was not the case for the farther inland regions (WA-C, NA-B and EA-B; middle rows in Figs. 5 – 7 ) where the equivalent Standard curves were generally below the confidence estimates of the reanalysis and best track derived curves. This indicates that the reanalysis wind gusts did not decrease as would be expected farther inland and as compared to the Standard in some cases, particularly in the coarser ERA5 data (Fig. 6 ). For the West coast regions (WA-D and WA-C), there is no clear change in the ARI time-slices covering 1990–2004 and 2005–2021 for ERA5 data (Fig. 6 ). However, BARRA (Fig. 5 ) and the adjusted best track record (Fig. 7 ) both showed some evidence of a slight downward trend in TC wind gust strength, including significant changes occurring in both these datasets. For the best-track record, the downward trend was consistent with the pre-1990 period, as these TC-gusts were similar in strength to the 1990–2004 period. For the North coast regions (NA-C and NA-B), changes in the ERA5 ARI time slices (Fig. 6 ) depend on how much legitimacy is given to the pre-1990 data, as the first (1979–1989) and third (2005–2021) time-slices were nearly identical in terms of their respective ARI curves. If the pre-1990 data is not considered, there is evidence of an upward trend in ERA5 wind gust strength [i.e., a significant increase of ~ 20 m s − 1 based on an ARI of 100] here in recent years, although this is much more likely related to decadal trends. Overall, BARRA showed that a slight downward trend in wind gust strength (~ 10 m s − 1 for an ARI of 100) may have occurred over this same time period, although the BARRA results were not significantly different between the two time-slices and became negligible when considering the farther inland region (NA-B). The best track results were once again more similar to BARRA than ERA5, although no significant trends in wind gust strength were noted between the two most recent ARI time-slices (Fig. 7 ). For best track data, the extended pre 1990 data more closely resembled the 1990–2004 period (at least for NA-C) in terms of TC wind gust strength. In contrast, the pre-1990 ERA5 gusts more closely resembled the later ERA5 period (2005–2021) TC-gust strength (Figs. 6 – 7 ). For the East coast regions (EA-C and EA-B), a slight upward trend, significant for EA-B, between the two later ARI time slices is noted based on the ERA5 data (Fig. 6 ). It is interesting to note that the two eastern regions are the only cases where the ERA5 and best-track data seem to agree on the historical trends in TC wind gust strength, showing quite similar representations of each time-slice. BARRA also indicated a slight increase in TC wind gust strength over the most recent time-slice for the EA-C region (~ 5 m s − 1 for an ARI of 100; Fig. 5 ), though this was not significant. Curiously, the reverse was true for the EA-B region (Fig. 5 ). This could potentially indicate that over the 1990–2004 period, winds maintained their strength farther inland than in the later period, though this was not noted in the best track data and could be due to a number of factors, including uncertainty in the BARRA data. 3.2.2 Subtropical regions For the two subtropical regions (denoted with the suffix “-S” in Fig. 1 ), ARIs are shown in the lower panels of Figs. 5 – 7 . EA-S, is the only region analyzed in this study that has not been designated an additional climate change multiplier in the latest AS/NZS Standard 1 . The analysis undertaken in this study appears to support that decision as both the TC and ETC curves from all datasets are well below the equivalent Standard curves (Figs. 5 – 7 , bottom row). Results between all three datasets are complementary in suggesting a potential small insignificant decrease for TC wind strength (Figs. 5 – 7 ); while the two reanalysis datasets are complementary in suggesting a small insignificant increase for ETC wind strength (Figs. 5 – 6 ). For the WA-S region, the TC curves also occur below the equivalent Standard curve with exception of the under-sampled pre-1990 best track curve (that is inconsistent with pre-1990 ERA5). Results between the two reanalysis datasets are complementary in suggesting a small but insignificant increase in wind strength (Figs. 5 – 6 ), although this was not the case for best track data (Fig. 7 ). Results for the ETC curves indicated little change and occurred comfortably below the design Standard (even below that of a “B” designation which a majority of the WA-S region poleward of 27°S is currently designated). 4. Discussion Wind speed ARIs determined from TC climate data in two reanalysis datasets, and best-track observations for context, were compared to AS/NZS wind Standard curves 1 . Results indicated moderate reproduction of the ARI metric with both reanalysis datasets, although comparisons deteriorated when analyzing wind speeds farther inland than 50 km. The finer resolution BARRA product better simulated the Standard's drop in TC-wind speed farther inland as compared to ERA5, an expected result. Differences in results between the three datasets highlight how important it is to consider multiple lines of evidence in these types of applications, especially with an eye to similar analyses using climate model simulations for future projections. Overall, in terms of providing new lines of evidence in relation to historical TC-wind gust strength experienced in coastal regions of Australia in recent decades, analysis here of multiple datasets provides some evidence of a likely (2/3 datasets) slight downtrend in observed and estimated tropical West coast TC-wind gust strength with low-to-medium confidence, consistent with a prior study 5 . Similarly, there is some evidence of a likely (2/3 datasets) slight uptrend in the East coast TC-wind gust data with low-to-medium confidence consistent with other studies 4,29 . Noting however it may be unrealistic to see actual trends related to warming over such a short time period. There was low consistency in results between different data sets on the existence of trends in TC-wind gust strength along the North coast of Australia. For the subtropical regions, according to the data analyzed here, a small insignificant uptrend in observed TC-wind speeds may have occurred since 1980 on the West coast (2/3 datasets), while a small insignificant downtrend may have occurred in observed TC-wind gust data on the East coast, while noting considerable uncertainties and large interannual variability makes it hard to determine this with a high degree of confidence. A significant trend in ETC-wind gust strength was found only for the subtropical East coast, where there was a small uptrend, consistent between both reanalysis datasets. West coast ETC wind gust strength was well below the current wind Standard, while East coast ETCs (East coast lows) were comparable, in terms of wind hazard, to the wind gust strength of TCs. Overall, it is difficult to attribute any changes in extreme wind gust strength in these datasets to anthropogenic climate change at this time, given the relatively short time period and limitations of the data quality including over longer time periods and prior to the era of widespread satellite coverage in this region. More confidence in attributing such changes to greenhouse warming could be achieved by applying similar methods to fine (regional scale) and very fine resolution (convection-permitting) climate model experiments under atmospheric forcing. While observations are often favored when evaluating trends, climate models have the strong advantage of having multiple ensemble members, meaning that physically driven changes might be discernible well before a single realization in the real world is detected. Indeed, work in this respect is underway by these authors on applying this method to several regional climate model (RCM) simulations that are of similar resolution or finer to the reanalysis data used here (e.g., a new version of BARRA is being produced that will downscale from ERA5 reanalysis at its host model, then downscale further again to convection-permitting scales). The work done here has spelt out limitations in what we would be expected to be encountered by similar application with climate model data. In addition, similar analysis focusing on identifying a potential evolution in extreme rainfall hazards (including that induced by cyclones) could also be highly beneficial and important for managing future climate risk in Australia. Declarations ACKNOWLEDGMENTS This work was supported by funding from the Earth Systems and Climate Change (ESCC) Hub of the Australian Government's National Environmental Science Programme (NESP). We would like to thank Acacia Pepler for supplying ETC tracks detected in ERA5 and helpful comments to improve the paper. We give further thanks to Geoffrey Boughton, John Holmes and Dörte Jakob whose expertise in this area greatly improved this work. Author contributions S.S.B., A.J.D., H.A.R., and S.S.C. conceived and designed the research. S.S.B. extracted the data and performed the analysis. The BARRA data was provided by C-H.S. and H.Y. S.S.B. wrote the manuscript with significant contributions from all co-authors. Additional Information The authors declare no competing interests. Data availability The best track tropical cyclone dataset as well as all data related to the ERA5 reanalysis are freely available online. The BARRA data is located on NCI Australia and is available to researchers on reasonable request. References Standards Australia. Structural design actions, Part 2: Wind actions, AS/NZS 1170.2:2021. Knutson, T. et al. Tropical Cyclones and Climate Change Assessment: Part II: Projected Response to Anthropogenic Warming. Bulletin of the American Meteorological Society, 101 , E303–E322, https://doi.org/10.1175/BAMS-D-18-0194.1 . (2020). Kossin, J. P., Knapp, K. R., Olander, T. L. & Velden, C. S. Global increase in major tropical cyclone exceedance probability over the past four decades. Proceedings of the National Academy of Sciences , 117 , 11975–11980, https://doi.org/10.1073/pnas.1920849117 . (2020). Holmes, J. D. Land-falling tropical cyclones on the queensland coast and implications of climate change for wind loads, Australian Journal of Structural Engineering, 21 , 135–142, https://doi.org/10.1080/13287982.2020.1717842 . (2020). Holmes, J. D. Tropical cyclone impacts on the Western Australian coast and extreme wind speeds in Region D. Australian Journal of Structural Engineering. 22 , 110–119, https://doi.org/10.1080/13287982.2021.1918376 . (2021). Su, C. H. et al. BARRA v1. 0: the bureau of meteorology atmospheric high-resolution regional reanalysis for Australia. Geoscientific Model Development, 12 , 2049–2068. (2019). Hersbach, H. et al. The ERA5 Global Reanalysis. Quarterly Journal of the Royal Meteorological Society, 146 , 1999–2049, https://doi.org/10.1002/qj.3803 . (2020). Murakami, H. Tropical cyclones in reanalysis data sets. Geophysical Research Letters, 41 , 2133–2141. (2014). Dvorak, V. F. A Technique for the Analysis and Forecasting of Tropical Cyclone Intensities From Satellite Pictures. NOAA. 5–8. (1973). Velden, C. et al. The Dvorak Tropical Cyclone Intensity Estimation Technique: A Satellite-Based Method That Has Endured For Over 30 Years. Bulletin of the American Meteorological Society, 87 , 1195–1214. (2006). Velden C, Burton, A. & Kuroiwa K. The First International Workshop on Satellite Analysis of Tropical Cyclones: Summary of Current Operational Methods to Estimate Intensity, Tropical Cyclone Research and Review , 1, 469–481, DOI: 10.6057/2012TCRR04.05 . (2012). Bell, S. S. et al. Australian tropical cyclone-induced extreme coastal winds in climate datasets. Bureau Research Report 056 , accessible online at http://www.bom.gov.au/research/publications/researchreports/BRR-056.pdf . (2021). Torn, R. D. & Snyder, C. Uncertainty of Tropical Cyclone Best-track Information. Weather and Forecasting, 27 , 715–729, https://doi.org/10.1175/WAF-D-11-00085.1 . (2012). Knaff, J. A., Brown, D. P., Courtney, J., Gallina, G. M. & Beven, J. L. An Evaluation of Dvorak Technique–Based Tropical Cyclone Intensity Estimates. Weather and Forecasting, 25 , 1362–1379, https://doi.org/10.1175/2010WAF2222375.1 . (2010). Courtney, J. B. et al. Revisions to the Australian tropical cyclone best track database. Journal of Southern Hemisphere Earth Systems Science, 71 , 203–227. (2021). Chand, S. S. et al. Review of tropical cyclones in the Australian region: Climatology, variability, predictability, and trends. WIREs Climate Change. 10 , e602. https://doi.org/10.1002/wcc.602 . (2019). Harper, B. A., Stroud, S. A., McCormack, M., & West, S. A review of historical tropical cyclone intensity in northwestern Australia and implications for climate change trends analysis. Australian Meteorological Magazine, 57 , 121–141. (2008). Klotzbach, P. J., Landsea, C. W. Extremely intense hurricanes: Revisiting Webster et al. (2005) after 10 years. Journal of Climate , 28 , 7621–7629, https://doi.org/10.1175/JCLI-D-15-0188.1 . (2015). Courtney, J. B & Burton, A. D. Joint Industry project for objective tropical cyclone reanalysis: final report, retrieved May 2 2021 at http://www.bom.gov.au/cyclone/history/database/OTCR-JIP_FinalReport_V1.3_public.pdf (2018). Courtney, J. B. et al. Towards an objective Tropical Cyclone dataset for the Australian region. Trop. Cyclone Res. 9 , 23–36. (2020). Paulsen, B. M. & Schroeder, J. L. An Examination of Tropical and Extratropical Gust Factors and the Associated Wind Speed Histograms. Journal of Applied Meteorology, 44 , 270–280, https://doi.org/10.1175/JAM2199.1 . (2005). Harper, B. A, Kepert, J. D., Ginger, J. D. Guidelines for converting between various wind averaging periods in tropical cyclone conditions. WMO , [available from https://www.wmo.int/pages/prog/www/tcp/documents/WMO_TD_1555_en.pdf] . (2010). Knapp, K. R., Kruk, M. C., Levinson, D. H., Diamond, H. J. & Neumann, C. J. The international best track archieve for climate stewardship (IBTrACS) unifying tropical cyclone data. Bulletin of the American Meteorological Society, 91 , 363–376. (2010). Su, C. H., Eizenberg, N. Known issues in the BARRA datasets. Bureau of Meteorology . Accessed April 7 2020 at http://www.bom.gov.au/research/projects/reanalysis/Known_issues_in_BARRA_dataset_July_2019.pdf . (2019). Simmonds, I. & Murray, R. J. Southern extratropical cyclone behavior in ECMWF analyses during the FROST Special Observing Periods. Weather and Forecasting, 14 , 878–891. (1999). Murray, R. J. & Simmonds, I. A numerical scheme for tracking cyclone centres from digital data Part I: development and operation of the scheme. Australian Meteorological Magazine, 39 , 155–166. (1991). Pepler, A. Record lack of cyclones in southern Australia during 2019. Geophysical Research Letters , 47 , e 2020GL088488 . https://doi.org/10.1029/2020GL088488. (2020). Hosking, J. R. M. L-Moments: Analysis and Estimation of Distributions Using Linear Combinations of Order Statistics. Journal of the Royal Statistical Society, 52 , 105–124. (1990). Wang, C., Wang, X., Khoo, Y. B. Extreme wind gust hazard in Australia and its sensitivity to climate change. Natural Hazards, 67 , 549–567, https://doi.org/10.1007/s11069-013-0582-5 . (2013). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Major revision 03 May, 2022 Reviews received at journal 18 Apr, 2022 Reviewers agreed at journal 11 Apr, 2022 Reviewers invited by journal 10 Apr, 2022 Editor assigned by journal 01 Apr, 2022 Editor invited by journal 31 Mar, 2022 Submission checks completed at journal 31 Mar, 2022 First submitted to journal 31 Mar, 2022 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-1507699","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":95017537,"identity":"c2a8fc4d-e255-4090-bc53-a059bced9f06","order_by":0,"name":"S. S. Bell","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYNCCAoYEBnbmxsckaDEAamFmbDYmWUubNFGK5dtPp0l8MGDI4wdqqS6o2MYg795jwPCzDY/5Z3K3Sc4wYCiWbGZsuz3jzG0GwzNnDBh78WlhyN0mzWPAkLjhMFALbxtQy4y0BAZePFrk+99uk/4D1LIfqKUYpoXxLx4tDDeAtjCAbAH6hRmkRV4i+QAzPlsMbrzdbNljIFEscZixWRroFx4DnsMHDsucw+ew3I03flTY5PG3Nx/8XFBxW06+vbHx4ZsyPA6DAAk4i8fgAAPDAYIaUO1tIE39KBgFo2AUDH8AACR4ThbzlkZBAAAAAElFTkSuQmCC","orcid":"","institution":"Bureau of Meteorology","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"S.","middleName":"S.","lastName":"Bell","suffix":""},{"id":95017541,"identity":"0abae6fd-d4b8-47c7-ac74-51cf9e451488","order_by":1,"name":"A. J. Dowdy","email":"","orcid":"","institution":"Bureau of Meteorology","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"A.","middleName":"J.","lastName":"Dowdy","suffix":""},{"id":95017543,"identity":"ffe6fd25-475b-4a35-b1fd-eb56249cb733","order_by":2,"name":"H. A. Ramsay","email":"","orcid":"","institution":"CSIRO Oceans and Atmosphere","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"H.","middleName":"A.","lastName":"Ramsay","suffix":""},{"id":95017544,"identity":"1f19850e-4d5d-4d90-8f7d-1e6ea531db9a","order_by":3,"name":"S. S. Chand","email":"","orcid":"","institution":"Federation University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"S.","middleName":"S.","lastName":"Chand","suffix":""},{"id":95017546,"identity":"053e751c-0e93-44b4-b4cf-815a6524153d","order_by":4,"name":"C-H Su","email":"","orcid":"","institution":"Bureau of Meteorology","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"C-H","middleName":"","lastName":"Su","suffix":""},{"id":95017548,"identity":"27c40b3e-9f1c-478a-99df-58e7791b31fd","order_by":5,"name":"H. Ye","email":"","orcid":"","institution":"Bureau of Meteorology","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"H.","middleName":"","lastName":"Ye","suffix":""}],"badges":[],"createdAt":"2022-03-31 04:14:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-1507699/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-1507699/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":19935190,"identity":"39727f56-0e3b-42bd-8488-7a48812407f2","added_by":"auto","created_at":"2022-04-04 14:43:57","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":122840,"visible":true,"origin":"","legend":"\u003cp\u003eWind speed regions used in this study. These regions have been designated by the historical wind strength the region has experienced (AS/NZS 1170.2:2021).\u0026nbsp;Considering the strongest designation in each of our regions, in descending order of strength these correspond to: Region D (WA-D), Region C (WA-C, WA-S, NA-C and EA-C) and Region B (NA-B, EA-B and EA-S).\u003c/p\u003e","description":"","filename":"fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/241aa5906ce9291442b29338.png"},{"id":19934455,"identity":"726eafdd-3f59-405c-a865-a7da81ccfdba","added_by":"auto","created_at":"2022-04-04 14:38:57","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":494410,"visible":true,"origin":"","legend":"\u003cp\u003eTime series and trend lines of calibrated BARRA reanalysis daily maximum wind gusts (3-second) occurring in each region. Daily maximum wind gusts are classified as “TC-related”, “ETC-related”, or “non-cyclone related” (see legend). Linear trend lines are shown for TC-day max winds (orange) and ETC-day max winds (green, where applicable). The estimate of the gradient (\u003cem\u003em\u003c/em\u003e)\u0026nbsp;and associated p-value (p) for these trend lines are shown above each time series. Note that the gradient are shown for years (rather than TC days) for interpretability.\u003c/p\u003e","description":"","filename":"fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/851f5b1371dd4e04777605cc.png"},{"id":19935191,"identity":"1bb3a1dd-3618-4619-846b-b48ea38e1897","added_by":"auto","created_at":"2022-04-04 14:43:57","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":448901,"visible":true,"origin":"","legend":"\u003cp\u003eAs in Fig. 2 but for the calibrated ERA5 reanalysis. As this data extends back to 1979, an additional trend line is fitted (black lines). Daily maximum wind gusts occurring before 1990 have been greyed out.\u0026nbsp;\u003c/p\u003e","description":"","filename":"fig3.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/95d2f21ebb4f1a171e486d3b.png"},{"id":19934450,"identity":"c4663907-f637-4042-a2f2-2595bb5528cd","added_by":"auto","created_at":"2022-04-04 14:38:57","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":337842,"visible":true,"origin":"","legend":"\u003cp\u003eAs in Fig. 2 but for best track TC wind gust speeds [replaced with adjusted V\u003csub\u003emax\u003c/sub\u003e over the period 1981-2016 (Courtney and Burton 2018)].\u003c/p\u003e","description":"","filename":"fig4.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/5a3fefdc779ca4ede9a850a1.png"},{"id":19936182,"identity":"d57d34c2-d515-47aa-8c19-f93a15786420","added_by":"auto","created_at":"2022-04-04 14:48:57","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":249492,"visible":true,"origin":"","legend":"\u003cp\u003eAverage recurrence intervals (ARIs) of TC-related daily maximum wind gusts (0.2 seconds) for the calibrated BARRA reanalysis.\u003c/p\u003e","description":"","filename":"fig5.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/bdd68f77526f1aef25f77cde.png"},{"id":19934453,"identity":"5a344b59-df9b-4193-8883-560285156228","added_by":"auto","created_at":"2022-04-04 14:38:57","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":295635,"visible":true,"origin":"","legend":"\u003cp\u003eAs in Fig. 5 but for the calibrated ERA5 reanalysis.\u003c/p\u003e","description":"","filename":"fig6.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/3104e4ba93b88761c1cfa14d.png"},{"id":19935188,"identity":"b4c35294-75a9-4f18-95ca-6888098ee4ea","added_by":"auto","created_at":"2022-04-04 14:43:57","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":209367,"visible":true,"origin":"","legend":"\u003cp\u003eAs in Fig. 5 but for best track TC wind gust data.\u003c/p\u003e","description":"","filename":"fig7.png","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/f7cdb9b8c693a85d928212ef.png"},{"id":19936201,"identity":"ec011c31-9a6b-4965-a90a-1520e4ca0808","added_by":"auto","created_at":"2022-04-04 14:49:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":640982,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-1507699/v1/e468ea6a-10bc-488e-add5-82044a0a031f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Using historical tropical cyclone climate datasets to examine wind speed recurrence for coastal Australia","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eExtreme wind speeds induced by tropical cyclones (TCs) can have considerable impacts on coastal communities and infrastructure situated in tropical Australia. In light of anthropogenic-induced climate change, it is important to investigate the potential evolution of TC-wind hazards for the benefit of future planning and disaster risk management in Australia.\u003c/p\u003e\n\u003cp\u003eThe expected average recurrence interval (ARI) of extreme wind speeds is a metric widely used by engineers and climate scientists alike to gauge the wind hazard for specific regions. In Australia, this is described by the \u0026ldquo;structural design actions\u0026rdquo; for winds\u003csup\u003e1\u003c/sup\u003e herein referred to as \u0026ldquo;the Standard\u0026rdquo;. Different regions of Australia are assigned various designations (A, B, C or D, in ascending order of strength) based on the historical wind strength (ARI) observed in that region. For example, Australia\u0026rsquo;s tropical and subtropical coastlines have much higher designations (B\u0026ndash;D) than the southern and inland regions (A) due to the increased exposure to TCs. Historically, appropriate regional wind speed designations have been determined by weather station surface anemometer measurements (corrected for terrain, topography and gust duration) and to a lesser extent damage surveys from past events (including in some cases back calculations of maximum wind speeds from failed simple structures such as road signs) and these methods remain very useful.\u003c/p\u003e\n\u003cp\u003eIn the latest edition of the AS/NZS structural design Standard (1170.2:2021), for the first time, a climate change multiplier (1.05) was included as mandatory to consider in some regions of Australia. This is consistent with a broad review and synthesis of available lines of evidence that indicate an increase in TC intensity with projected future warming is more likely than a decrease\u003csup\u003e2,3\u003c/sup\u003e. For local context, the East coast of Australia has experienced a recent influx of severe Australian Category 4/5 TCs\u003csup\u003e4\u003c/sup\u003e (Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e) while the West coast of Australia has experienced fewer Category 4/5 TCs recently\u003csup\u003e5\u003c/sup\u003e. This leads to the question of what role can climate model experiments play in informing us of likely future changes in fine-scale metrics such as wind speed ARIs that have traditionally come from observed ground truth data. Climate models, despite having their own limitations such as biases and inaccurate representations of some small- and large-scale physical processes, still provide homogenous simulations that can be useful for examining the potential influence of climate change now and into the future, as a complementary line of evidence to observations-based data (which also have limitations including due to evolving operational practices for wind observations in Australia with a relatively short time period of high-quality data being available).\u003c/p\u003e\n\u003ctable border=\"1\" id=\"Tab1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThe Australian TC wind scale as defined by wind gust speed (3-second).\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCategory\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGust speed\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e24.5\u0026ndash;34 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e34\u0026ndash;45.5 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e45.5\u0026ndash;61.6 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e61.6\u0026ndash;77.4 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;77.4 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eIt is timely that with the recent release of new, finer resolution reanalysis products, including the Bureau of Meteorology Atmospheric high-resolution Regional Reanalysis for Australia (BARRA\u003csup\u003e6\u003c/sup\u003e) and the global ERA5 reanalysis\u003csup\u003e7\u003c/sup\u003e produced by the ECMWF; the feasibility of determining wind speed ARIs in climate model type products can be tested. This can be achieved by direct comparison with the ground truth determined wind speed ARIs from AS/NZS 1170.2. The two reanalysis datasets can also provide additional lines of evidence in relation to current historical TC trends and extreme wind speeds affecting Australia.\u003c/p\u003e\n\u003cp\u003eThe simulation of peak wind gust speeds from TCs in reanalysis products poses many challenges, and the products used here do not completely resolve the absolute wind gust speeds of the more intense TCs\u003csup\u003e2,8\u003c/sup\u003e. However, by calibrating the reanalysis data to the observations, such as from Automatic Weather Stations (AWS) and from the best track estimates of TC wind speeds\u003csup\u003e9,10,11\u003c/sup\u003e useful information on absolute gust speeds in reanalysis data can be obtained, though with some obvious limitations\u003csup\u003e12\u003c/sup\u003e. That is, we are not suggesting that the ARIs calculated here should replace the current structural AS/NZS wind speed Standard, but rather to serve as an additional line of evidence contributing to a more comprehensive understanding of potential long-term changes in this metric and in the risk of TC-wind hazards for coastal Australia. In particular, several data sets are used here for the first time to examine TCs and associated extreme wind speeds in these regions, thereby providing new insight on this important topic, which is intended to help provide useful guidance complementary to existing information.\u003c/p\u003e\n\u003cp\u003eHere, we evaluate ARIs of TC-related daily-maximum wind gusts over eight Australian coastal regions (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). These regions closely correspond with those designated by AS/NZS 1170.2:2021, allowing for a comparison of these wind speeds to the Standard. Two subtropical regions on either side of the Australian continent, where weakening TCs and extratropical cyclones (ETCs) can sometimes occur, are additionally examined. Examination of these two regions may provide some guidance on a key scientific question around the potential poleward movement of TCs and their wind hazards. The following section provides details on the Data and Methods used. Results are shown in Section 3, and a Summary and Conclusion is provided in Section 4.\u003c/p\u003e"},{"header":"2. Data And Methods","content":"\u003cdiv class=\"Section2\" id=\"Sec3\"\u003e\n \u003ch2\u003e2.1. Data\u003c/h2\u003e\n \u003cdiv class=\"Section3\" id=\"Sec4\"\u003e\n \u003ch2\u003e2.1.1 Best track\u003c/h2\u003e\n \u003cp\u003eThe Bureau of Meteorology (BoM) TC dataset (available online at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bom.gov.au/cyclone/tropical-cyclone-knowledge-centre/databases/\u003c/span\u003e\u003c/span\u003e) includes historical TCs that have occurred in the Australian region (from 90 to 160 degrees East). Since the introduction of operational satellite technology in the early 1970s, the historical records of TCs (analyzed here over the period 1971 to 2021), have included best track sustained wind speed, wind gust estimates, and TC position and time information (with a degree of uncertainty\u003csup\u003e13\u003c/sup\u003e). We note that some TCs in the most recent two years of the dataset are considered provisional, meaning they have not yet undergone formal best track review; and that several TCs in the early 1970s do not have wind speed information available.\u003c/p\u003e\n \u003cp\u003eTC wind speed estimates in the best track data have been most widely determined by the satellite-based Dvorak technique\u003csup\u003e9,10,14\u003c/sup\u003e. This technique involves the subjective assignment of Dvorak numbers to satellite imagery of TCs to give an indication of the wind speed. However, especially more recently, other tools and techniques have been used by operational forecast centres and meteorological agencies to estimate TC intensities. Such tools and techniques, which include scatterometers, AWS, as well as the Advanced Dvorak technique (ADT), and the SATellite CONsensus intensity approach (SATCON), can influence the final Dvorak number estimate (and therefore the intensity) that is eventually recorded in the best track data\u003csup\u003e11\u003c/sup\u003e. These new tools and techniques make the recent (~\u0026thinsp;post 2003) best-track record more accurate\u003csup\u003e15\u003c/sup\u003e, but also embeds a degree of inhomogeneity in the dataset when compared with the earlier periods. As such, numerous studies have recommended data from more recent periods for evaluations of TC intensity estimates\u003csup\u003e16\u003c/sup\u003e. For example, [\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e] recommends data from 1980/81 onwards for TC intensity analysis, whereas others are more conservative: recommending data from 1990 onwards\u003csup\u003e18\u003c/sup\u003e and from 2003 onwards\u003csup\u003e19\u003c/sup\u003e. In our case, we have used the entire satellite data for completeness, but placed emphasis on the period from 1990 onwards.\u003c/p\u003e\n \u003cp\u003eA re-evaluation of TCs in the Australian region\u003csup\u003e19,20\u003c/sup\u003e provides revised wind speed (V\u003csub\u003emax\u003c/sub\u003e) estimates in some instances compared to those in the best-track (BoM) TC dataset over the period 1981\u0026ndash;2016, by for example, using ADT to remove subjectivity from the simple Dvorak technique. In addition, there are some cases in the best track dataset where only a TC\u0026rsquo;s sustained wind speed (i.e., 10-minute average value) was recorded in the best track TC dataset (i.e., no wind gust conversion based on an open terrain estimate is provided). Here, 3-second wind gusts were obtained by multiplying the wind speed value by a gust factor (GF) of 1.4. GFs can vary due to several factors\u003csup\u003e21\u003c/sup\u003e, including changes in surface roughness (where complex terrain requires a higher GF than that of the smooth open ocean\u003csup\u003e22\u003c/sup\u003e). Similarly, there are also various modification factors that determine the actual wind speed a specific site will experience. These factors include the directional, shielding, terrain and topographic multipliers (e.g., AS/NZS 1170.2). Due to the broad scale of the analyses undertaken here, a GF of 1.4 was used when necessary, noting a similar value was used in most best track derivations of wind gust speed (that vary depending on terrain).\u003c/p\u003e\n \u003cp\u003eAs discussed in the previous paragraph, there are three options to use for representing best track wind gust values. We give a preference to the Courtney and Burton\u003csup\u003e20\u003c/sup\u003e adjusted V\u003csub\u003emax\u003c/sub\u003e (10-min sustained) wind speeds, where available, ahead of the best-track wind speeds such that the final TC wind gust speed values assigned to TC observations are in the following order:\u003c/p\u003e\n\u003cp\u003e1. Adjusted V\u003csub\u003emax\u003c/sub\u003e \u003cstrong\u003e✕\u003c/strong\u003e 1.4 (Courtney and Burton\u003csup\u003e20\u003c/sup\u003e) (Predominantly occurring from 1980\u0026ndash;2016)\u003c/p\u003e \u003cspan\u003e\n \u003cp\u003e2. Best track wind gust speed estimate (Predominantly occurring everywhere else)\u003c/p\u003e\n \u003c/span\u003e \u003cspan\u003e\n \u003cp\u003e3. Best track V\u003csub\u003emax\u003c/sub\u003e \u003cstrong\u003e✕\u003c/strong\u003e 1.4 (occur infrequently preceding 1980)\u003c/p\u003e\n \u003c/span\u003e\n\u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec5\"\u003e\n \u003ch2\u003e2.1.2 ERA5\u003c/h2\u003e\n \u003cp\u003eThe latest global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF), ERA5\u003csup\u003e7\u003c/sup\u003e, is available at the horizontal resolution of 31 km. Daily maximum wind gusts (averaged over 3-seconds at 10 m height) were calculated from hourly maximum values over the available period spanning 1979 to June 2021. Note that the effects of convection are only included in the wind gust parameterization from October 2008 onwards which embeds a degree of non-homogeneity in the data as is the case in general for data based on observations.\u003c/p\u003e\n \u003cp\u003eFor comparison with the Standard, the ERA5 wind speeds are calibrated to observational data (mostly derived over the ocean) using quantile-quantile scaling\u003csup\u003e12\u003c/sup\u003e. These winds are then further converted to standard terrain conditions (i.e., over land) by a factor of 1.05 [\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e]. For ARI calculations, the daily-maximum wind gusts are additionally converted to a 0.2 second averaging period (1.1), with exception of the tail 1% of quantiles. This was done to avoid exacerbating the already large absolute values of the distribution tail (due to the relatively large calibration applied on this data). Combining these conversions, the calibration factors for the original ERA5 wind data range from 1.1 (for the lowest intensity gusts) to a maximum of 1.85 for raw wind gusts\u0026thinsp;\u0026gt;\u0026thinsp;45 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec6\"\u003e\n \u003ch2\u003e2.1.3 BARRA\u003c/h2\u003e\n \u003cp\u003eBARRA, is a 12 km horizontal grid spacing reanalysis product covering the Australian region, (65\u0026deg;E \u0026ndash; 163\u0026deg;W, 20\u0026deg;N \u0026ndash; 60\u0026deg;S). This reanalysis includes assimilation of central pressure observations from the International Best Track Archive for Climate Stewardship (IBTrACS\u003csup\u003e23\u003c/sup\u003e). Maximum hourly surface (10 m) wind gust speeds were acquired and used to calculate daily maximum wind gust values over the available period spanning 1990 to February 2019.\u003c/p\u003e\n \u003cp\u003eThere are several known issues in the BARRA data\u003csup\u003e24\u003c/sup\u003e, including grid-sized storms and anomalous wind gust spikes that can produce exceedingly large wind gust values at single isolated grid-points, particularly over northern Australia and near coastlines. To circumvent this issue, the hourly maximum grid cell for a specific region is discarded and replaced by the second-maximum value. The hour-on-hour increase in maximum wind gust speed is also limited to 20 ms\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the final calculation of daily-maximum wind gust speeds. In cases where there is no grid-point storm or extreme gust spike present, this method usually results in a negligible reduction (\u0026lt;\u0026thinsp;1 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) in the value of the maximum wind gust speed.\u003c/p\u003e\n \u003cp\u003eAs was done for the ERA5 data, the BARRA reanalysis data are calibrated using quantile-quantile scaling to match the observational data while preserving the original distribution of the data\u003csup\u003e12\u003c/sup\u003e. Slightly different to the ERA5 data, the entire BARRA distribution is converted to an averaging period of 0.2 seconds (1.1) and a larger value is used\u003csup\u003e1,12\u003c/sup\u003e (1.13) to convert to standard terrain conditions. As compared to the calibration of ERA5 data, the finer resolution BARRA data required smaller calibration factors (with a maximum calibration factor of 1.35 for raw wind gusts\u0026thinsp;\u0026gt;\u0026thinsp;70 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), such that this approach provides a suitable means of reducing limitations of the reanalysis products while producing calibrated data that are more consistent with the observations in absolute terms.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Section2\" id=\"Sec7\"\u003e\n \u003ch2\u003e2.2. Methods\u003c/h2\u003e\n \u003cdiv class=\"Section3\" id=\"Sec8\"\u003e\n \u003ch2\u003e2.2.1 Wind speed regions\u003c/h2\u003e\n \u003cp\u003eEight wind speed regions (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e) are defined here, where the tropical regions are denoted by their AZ/NZS 1170.2 designation \u0026ldquo;-B\u0026rdquo;, \u0026ldquo;-C\u0026rdquo; or \u0026ldquo;-D\u0026rdquo; suffix. The two subtropical regions are simply denoted by the \u0026ldquo;-S\u0026rdquo; suffix. The strongest designation in each region corresponds to the regions used here uniformly with exception of the WA-S region which lies only partly under the Region-C designation (between 25\u0026ndash;27\u0026deg;S). In AS/NZS 1170.2:2021, all regions with exception of EA-S have a mandatory climate change multiplier (1.05) included in their ARI calculations. However, as this study focuses on the present and historical climate, these climate change multipliers are not considered here.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec9\"\u003e\n \u003ch2\u003e2.2.2 TC-related wind gusts\u003c/h2\u003e\n \u003cp\u003eMeasures of wind gust maxima are confined to the eight wind speed regions as shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. Daily maximum wind gusts for these regions are considered to be \u0026ldquo;TC-related\u0026rdquo; if there is a TC track position either inside the region, or its centre is within a 4\u003csup\u003e\u0026deg;\u003c/sup\u003e radius on the same day as the daily-maximum wind gust. Such \u0026ldquo;days\u0026rdquo; that are determined to be TC-related are referred to as \u0026ldquo;TC days\u0026rdquo;. The TC data is defined from the BoM TC dataset (best track) and extra-tropical transition days as flagged in the database were not removed from the analysis.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec10\"\u003e\n \u003ch2\u003e2.2.3 Detection of ETC tracks and ETC-related wind gusts\u003c/h2\u003e\n \u003cp\u003eFor the two subtropical regions denoted with the suffix \u0026ldquo;-S\u0026rdquo; in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, extratropical cyclone (ETC) days are also considered, in addition to TC days (i.e., days on which ETCs occur inside or within 4\u003csup\u003e\u0026deg;\u003c/sup\u003e radius of a given region). The cyclone tracking scheme of Murray and Simmonds\u003csup\u003e25\u003c/sup\u003e with refinements\u003csup\u003e26\u003c/sup\u003e was used to detect their occurrence in ERA5 data over the full time period and applied to both ERA5 and BARRA wind gust data. This scheme has been applied previously in reanalysis data to produce ETC tracks\u003csup\u003e27\u003c/sup\u003e. The number of ETCs was not tuned to any existing climatological records, and so lower intensity circulations can remain in our definition of \u0026ldquo;ETC days\u0026rdquo;. Detected ETCs were differentiated from other storms, such as detected TCs, by having a genesis location poleward of 23\u003csup\u003e\u0026deg;\u003c/sup\u003eS.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec11\"\u003e\n \u003ch2\u003e2.2.4 Extreme value modelling, ARIs and peaks\u003c/h2\u003e\n \u003cp\u003eThe method of peak-over-threshold was used to model extreme daily-maximum cyclone-related wind gust data, where the threshold was arbitrarily chosen to produce a total average of ~\u0026thinsp;3 peaks per year. In cases where there are insufficient peaks for a set of data/region (e.g., subtropical regions and best track data) the data is resampled with replacement to achieve the desired number of peaks. This procedure is repeated several times to ensure the most extreme values are not oversampled. Confidence bounds for ARI curves are derived by removing (lower bound) and duplicating (upper bound) the most extreme values (top 5%).\u003c/p\u003e\n \u003cp\u003eThe lognormal and Weibull functions best modelled the tails of our wind gust distributions in comparison to other functions (such as GEV and Gumbel; not shown). It should be noted that the Standard curves are derived using GEV, while curves derived from our data are more asymptotic (based on lognormal and Weibull shapes). The parameter estimation method of maximum likelihood was preferred although others could have been used (e.g., L-moments\u003csup\u003e28\u003c/sup\u003e). In cases where the lognormal function did not provide a good fit to our data (mostly for cases with tail-heavy distributions on the West coast and for best track data), the Weibull distribution was used instead.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv class=\"Section2\" id=\"Sec13\"\u003e\n \u003ch2\u003e3.1. Time series\u003c/h2\u003e\n \u003cp\u003eTime series of daily-maximum wind gusts (3-seconds) related to TCs, \u0026ldquo;non-cyclones\u0026rdquo; and \u0026ldquo;ETCs\u0026rdquo; (for the two subtropical regions) are shown for three datasets: BARRA (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e) ERA5 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e) and Best track data (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\n \u003cdiv class=\"Section3\" id=\"Sec14\"\u003e\n \u003ch2\u003e3.1.1 Tropical cyclone days\u003c/h2\u003e\n \u003cp\u003eLinear trend lines are fitted to each time series of TC day maximum wind gusts (Figs. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e) and none are found to be significant at the 95% confidence level (i.e., p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), with exception of best track data back to 1971 in the EA-S region (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). However, the limited number of samples and large differences between trendlines beginning from different start points (1990 and 1971), embeds a lot of doubt in this result.\u003c/p\u003e\n \u003cp\u003eCollectively, there are also differences in sign between at least one of the three datasets in each region. For example, if we consider that the different datasets provide a confidence range of the trendline coefficients for the combined tropical regions since 1990, these can be interpretated as:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts on the Western Australian coast are insignificantly changing at a rate of [-0.5\u003csup\u003e*\u003c/sup\u003e (-0.01) to 0.06] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts on the Northern Australian coast are insignificantly changing at a rate of [-0.2\u003csup\u003e*\u003c/sup\u003e (-0.06) to 0.1] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts on the Eastern Australian coast are insignificantly changing at a rate of [-0.08 to (0.07) 0.5\u003csup\u003e*\u003c/sup\u003e] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cp\u003eExcessively large annual trends (\u0026gt;\u0026thinsp;0.2 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e per year) all came from the best track dataset and are very likely influenced by inhomogeneities in that data. For these cases, the second largest annual trend from either reanalysis dataset are shown beside the original trends in curly brackets.\u003c/p\u003e\n \u003cp\u003eFor the two subtropical regions since 1990:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts affecting the EA-S region have a consistent downtrend between datasets but are insignificantly changing at a rate of [-0.25 to -0.1] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts affecting the WA-S region are insignificantly changing at a rate of [-0.07 to 0.15] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cp\u003eWhen considering trendlines back to the 1970s (i.e., from ERA5 and best track data; Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e), the trendline coefficients often become smaller, with exception of the tropical West coast regions and the subtropical regions for best track data (where trendline coefficients often become larger in size).\u003c/p\u003e\n \u003cp\u003eIt is also practical to consider trends in the more extreme wind gusts that are most likely to cause significant property damage (i.e., Category 3 or above; \u0026gt;45 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). Trendlines for these \u0026ldquo;extreme\u0026rdquo; TC wind gusts (not shown) in combined regions since 1990 from the two reanalysis datasets can be interpreted as:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts on the Western Australian coast are in some cases significantly changing at a rate of [-0.53 to 0.06] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts on the Northern Australian coast are insignificantly changing at a rate of [-0.22 to 0.14] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eTC day maximum wind gusts on the Eastern Australian coast are insignificantly changing at a rate of [-0.25 to 0.12] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec15\"\u003e\n \u003ch2\u003e3.1.2 Extratropical cyclone days\u003c/h2\u003e\n \u003cp\u003eConsidering the ETC day trend lines that are fitted to the two subtropical regions (BARRA and ERA5 only; Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e) the only significant result at the 95% confidence level (i.e., p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) was that of ETC wind gusts affecting the EA-S region since 1990 in ERA5: indicating a rate of increase of 0.2 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e per year. This result was consistent in sign but was much greater in size when compared with BARRA and additionally ERA5 data back to 1979. There was no trend in ETC wind gust strength affecting the WA-S region.\u003c/p\u003e\n \u003cp\u003eFor completeness, since 1990:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eETC day maximum wind gusts affecting EA-S are in one case significantly changing at a rate of [-0.1 to 0.2] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eETC day maximum wind gusts affecting WA-S are insignificantly changing at a rate of [0 to 0] m s\u003csup\u003e-1\u003c/sup\u003e per year.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Section2\" id=\"Sec16\"\u003e\n \u003ch2\u003e3.2. ARIs\u003c/h2\u003e\n \u003cp\u003eFor comparison with the Standard wind curves, observed and model data are converted to an averaging period of 0.2 second gusts. However, this is still not a fair comparison as the Standard curves are extrapolations based on hundreds of years of data at several point locations inside a region while the reanalysis and best track curves are based on fewer years of data but at every available location (i.e., every grid cell in a region). Additionally, the reanalysis data have been calibrated to represent absolute wind speeds up to Category 5. So, it is expected that the curves based on the calibrated reanalyses (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e) and best track data (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e) would have higher returns for small-to-medium ARIs and then approach the Standard when ARI is large (\u0026gt;\u0026thinsp;500). In an attempt to provide a fairer comparison between our ARI curves and the Standard ARI curves, we divide the Standard ARI curves by 33 (i.e., 500-year ARIs become 15-year and so on) to get equivalent curves for comparison which are additionally displayed to the regular Standard ARIs (dashed and solid black curves in Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\n \u003cdiv class=\"Section3\" id=\"Sec17\"\u003e\n \u003ch2\u003e3.2.1 Tropical regions\u003c/h2\u003e\n \u003cp\u003eFor the regions nearest to coastlines (WA-D, NA-C and EA-C; top rows in Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e) the equivalent Standard curve generally lies in between or within the confidence estimates of the reanalysis and best track derived curves. However, this was not the case for the farther inland regions (WA-C, NA-B and EA-B; middle rows in Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e) where the equivalent Standard curves were generally below the confidence estimates of the reanalysis and best track derived curves. This indicates that the reanalysis wind gusts did not decrease as would be expected farther inland and as compared to the Standard in some cases, particularly in the coarser ERA5 data (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eFor the West coast regions (WA-D and WA-C), there is no clear change in the ARI time-slices covering 1990\u0026ndash;2004 and 2005\u0026ndash;2021 for ERA5 data (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e). However, BARRA (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e) and the adjusted best track record (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e) both showed some evidence of a slight downward trend in TC wind gust strength, including significant changes occurring in both these datasets. For the best-track record, the downward trend was consistent with the pre-1990 period, as these TC-gusts were similar in strength to the 1990\u0026ndash;2004 period.\u003c/p\u003e\n \u003cp\u003eFor the North coast regions (NA-C and NA-B), changes in the ERA5 ARI time slices (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e) depend on how much legitimacy is given to the pre-1990 data, as the first (1979\u0026ndash;1989) and third (2005\u0026ndash;2021) time-slices were nearly identical in terms of their respective ARI curves. If the pre-1990 data is not considered, there is evidence of an upward trend in ERA5 wind gust strength [i.e., a significant increase of ~\u0026thinsp;20 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e based on an ARI of 100] here in recent years, although this is much more likely related to decadal trends. Overall, BARRA showed that a slight downward trend in wind gust strength (~\u0026thinsp;10 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003efor an ARI of 100) may have occurred over this same time period, although the BARRA results were not significantly different between the two time-slices and became negligible when considering the farther inland region (NA-B). The best track results were once again more similar to BARRA than ERA5, although no significant trends in wind gust strength were noted between the two most recent ARI time-slices (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e). For best track data, the extended pre 1990 data more closely resembled the 1990\u0026ndash;2004 period (at least for NA-C) in terms of TC wind gust strength. In contrast, the pre-1990 ERA5 gusts more closely resembled the later ERA5 period (2005\u0026ndash;2021) TC-gust strength (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eFor the East coast regions (EA-C and EA-B), a slight upward trend, significant for EA-B, between the two later ARI time slices is noted based on the ERA5 data (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e). It is interesting to note that the two eastern regions are the only cases where the ERA5 and best-track data seem to agree on the historical trends in TC wind gust strength, showing quite similar representations of each time-slice. BARRA also indicated a slight increase in TC wind gust strength over the most recent time-slice for the EA-C region (~\u0026thinsp;5 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for an ARI of 100; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e), though this was not significant. Curiously, the reverse was true for the EA-B region (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). This could potentially indicate that over the 1990\u0026ndash;2004 period, winds maintained their strength farther inland than in the later period, though this was not noted in the best track data and could be due to a number of factors, including uncertainty in the BARRA data.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Section3\" id=\"Sec18\"\u003e\n \u003ch2\u003e3.2.2 Subtropical regions\u003c/h2\u003e\n \u003cp\u003eFor the two subtropical regions (denoted with the suffix \u0026ldquo;-S\u0026rdquo; in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e), ARIs are shown in the lower panels of Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e\n \u003cp\u003eEA-S, is the only region analyzed in this study that has not been designated an additional climate change multiplier in the latest AS/NZS Standard\u003csup\u003e1\u003c/sup\u003e. The analysis undertaken in this study appears to support that decision as both the TC and ETC curves from all datasets are well below the equivalent Standard curves (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e, bottom row). Results between all three datasets are complementary in suggesting a potential small insignificant decrease for TC wind strength (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e); while the two reanalysis datasets are complementary in suggesting a small insignificant increase for ETC wind strength (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eFor the WA-S region, the TC curves also occur below the equivalent Standard curve with exception of the under-sampled pre-1990 best track curve (that is inconsistent with pre-1990 ERA5). Results between the two reanalysis datasets are complementary in suggesting a small but insignificant increase in wind strength (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e), although this was not the case for best track data (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e). Results for the ETC curves indicated little change and occurred comfortably below the design Standard (even below that of a \u0026ldquo;B\u0026rdquo; designation which a majority of the WA-S region poleward of 27\u0026deg;S is currently designated).\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eWind speed ARIs determined from TC climate data in two reanalysis datasets, and best-track observations for context, were compared to AS/NZS wind Standard curves\u003csup\u003e1\u003c/sup\u003e. Results indicated moderate reproduction of the ARI metric with both reanalysis datasets, although comparisons deteriorated when analyzing wind speeds farther inland than 50 km. The finer resolution BARRA product better simulated the Standard's drop in TC-wind speed farther inland as compared to ERA5, an expected result. Differences in results between the three datasets highlight how important it is to consider multiple lines of evidence in these types of applications, especially with an eye to similar analyses using climate model simulations for future projections.\u003c/p\u003e \u003cp\u003eOverall, in terms of providing new lines of evidence in relation to historical TC-wind gust strength experienced in coastal regions of Australia in recent decades, analysis here of multiple datasets provides some evidence of a likely (2/3 datasets) slight downtrend in observed and estimated tropical West coast TC-wind gust strength with low-to-medium confidence, consistent with a prior study\u003csup\u003e5\u003c/sup\u003e. Similarly, there is some evidence of a likely (2/3 datasets) slight uptrend in the East coast TC-wind gust data with low-to-medium confidence consistent with other studies\u003csup\u003e4,29\u003c/sup\u003e. Noting however it may be unrealistic to see actual trends related to warming over such a short time period. There was low consistency in results between different data sets on the existence of trends in TC-wind gust strength along the North coast of Australia.\u003c/p\u003e \u003cp\u003eFor the subtropical regions, according to the data analyzed here, a small insignificant uptrend in observed TC-wind speeds may have occurred since 1980 on the West coast (2/3 datasets), while a small insignificant downtrend may have occurred in observed TC-wind gust data on the East coast, while noting considerable uncertainties and large interannual variability makes it hard to determine this with a high degree of confidence. A significant trend in ETC-wind gust strength was found only for the subtropical East coast, where there was a small uptrend, consistent between both reanalysis datasets. West coast ETC wind gust strength was well below the current wind Standard, while East coast ETCs (East coast lows) were comparable, in terms of wind hazard, to the wind gust strength of TCs.\u003c/p\u003e \u003cp\u003eOverall, it is difficult to attribute any changes in extreme wind gust strength in these datasets to anthropogenic climate change at this time, given the relatively short time period and limitations of the data quality including over longer time periods and prior to the era of widespread satellite coverage in this region. More confidence in attributing such changes to greenhouse warming could be achieved by applying similar methods to fine (regional scale) and very fine resolution (convection-permitting) climate model experiments under atmospheric forcing. While observations are often favored when evaluating trends, climate models have the strong advantage of having multiple ensemble members, meaning that physically driven changes might be discernible well before a single realization in the real world is detected. Indeed, work in this respect is underway by these authors on applying this method to several regional climate model (RCM) simulations that are of similar resolution or finer to the reanalysis data used here (e.g., a new version of BARRA is being produced that will downscale from ERA5 reanalysis at its host model, then downscale further again to convection-permitting scales). The work done here has spelt out limitations in what we would be expected to be encountered by similar application with climate model data. In addition, similar analysis focusing on identifying a potential evolution in extreme rainfall hazards (including that induced by cyclones) could also be highly beneficial and important for managing future climate risk in Australia.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eACKNOWLEDGMENTS\u003c/p\u003e\n\u003cp\u003eThis work was supported by funding from the Earth Systems and Climate Change (ESCC) Hub of the Australian Government\u0026apos;s National Environmental Science Programme (NESP). We would like to thank Acacia Pepler for supplying ETC tracks detected in ERA5 and helpful comments to improve the paper. We give further thanks to Geoffrey Boughton, John Holmes and\u0026nbsp;D\u0026ouml;rte\u0026nbsp;Jakob\u0026nbsp;whose expertise in this area greatly improved this work.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eS.S.B., A.J.D., H.A.R., and S.S.C. conceived and designed the research. S.S.B. extracted the data and performed the analysis. The BARRA data was provided by C-H.S. and H.Y. S.S.B. wrote the manuscript with significant contributions from all co-authors.\u003c/p\u003e\n\u003cp\u003eAdditional Information\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\u003cp\u003eData availability\u003c/p\u003e\n\u003cp\u003eThe best track tropical cyclone dataset as well as all data related to the ERA5 reanalysis are freely available online. The BARRA data is located on NCI Australia and is available to researchers on reasonable request.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eStandards Australia. Structural design actions, Part 2: Wind actions, AS/NZS 1170.2:2021.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnutson, T. et al. Tropical Cyclones and Climate Change Assessment: Part II: Projected Response to Anthropogenic Warming. Bulletin of the American Meteorological Society, \u003cb\u003e101\u003c/b\u003e, E303\u0026ndash;E322, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/BAMS-D-18-0194.1\u003c/span\u003e\u003cspan address=\"10.1175/BAMS-D-18-0194.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKossin, J. P., Knapp, K. R., Olander, T. L. \u0026amp; Velden, C. S. Global increase in major tropical cyclone exceedance probability over the past four decades. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cb\u003e117\u003c/b\u003e, 11975\u0026ndash;11980, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.1920849117\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1920849117\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolmes, J. D. Land-falling tropical cyclones on the queensland coast and implications of climate change for wind loads, Australian Journal of Structural Engineering, \u003cb\u003e21\u003c/b\u003e, 135\u0026ndash;142, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/13287982.2020.1717842\u003c/span\u003e\u003cspan address=\"10.1080/13287982.2020.1717842\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolmes, J. D. Tropical cyclone impacts on the Western Australian coast and extreme wind speeds in Region D. Australian Journal of Structural Engineering. \u003cb\u003e22\u003c/b\u003e, 110\u0026ndash;119, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/13287982.2021.1918376\u003c/span\u003e\u003cspan address=\"10.1080/13287982.2021.1918376\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu, C. H. et al. BARRA v1. 0: the bureau of meteorology atmospheric high-resolution regional reanalysis for Australia. Geoscientific Model Development, \u003cb\u003e12\u003c/b\u003e, 2049\u0026ndash;2068. (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHersbach, H. et al. The ERA5 Global Reanalysis. Quarterly Journal of the Royal Meteorological Society, \u003cb\u003e146\u003c/b\u003e, 1999\u0026ndash;2049, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/qj.3803\u003c/span\u003e\u003cspan address=\"10.1002/qj.3803\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurakami, H. Tropical cyclones in reanalysis data sets. Geophysical Research Letters, \u003cb\u003e41\u003c/b\u003e, 2133\u0026ndash;2141. (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDvorak, V. F. A Technique for the Analysis and Forecasting of Tropical Cyclone Intensities From Satellite Pictures. NOAA. 5\u0026ndash;8. (1973).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVelden, C. et al. The Dvorak Tropical Cyclone Intensity Estimation Technique: A Satellite-Based Method That Has Endured For Over 30 Years. Bulletin of the American Meteorological Society, \u003cb\u003e87\u003c/b\u003e, 1195\u0026ndash;1214. (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVelden C, Burton, A. \u0026amp; Kuroiwa K. The First International Workshop on Satellite Analysis of Tropical Cyclones: Summary of Current Operational Methods to Estimate Intensity, \u003cem\u003eTropical Cyclone Research and Review\u003c/em\u003e, 1, 469\u0026ndash;481, DOI:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.6057/2012TCRR04.05\u003c/span\u003e\u003cspan address=\"10.6057/2012TCRR04.05\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBell, S. S. et al. Australian tropical cyclone-induced extreme coastal winds in climate datasets. \u003cem\u003eBureau Research Report 056\u003c/em\u003e, accessible online at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bom.gov.au/research/publications/researchreports/BRR-056.pdf\u003c/span\u003e\u003cspan address=\"http://www.bom.gov.au/research/publications/researchreports/BRR-056.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTorn, R. D. \u0026amp; Snyder, C. Uncertainty of Tropical Cyclone Best-track Information. Weather and Forecasting, \u003cb\u003e27\u003c/b\u003e, 715\u0026ndash;729, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/WAF-D-11-00085.1\u003c/span\u003e\u003cspan address=\"10.1175/WAF-D-11-00085.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnaff, J. A., Brown, D. P., Courtney, J., Gallina, G. M. \u0026amp; Beven, J. L. An Evaluation of Dvorak Technique\u0026ndash;Based Tropical Cyclone Intensity Estimates. Weather and Forecasting, \u003cb\u003e25\u003c/b\u003e, 1362\u0026ndash;1379, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/2010WAF2222375.1\u003c/span\u003e\u003cspan address=\"10.1175/2010WAF2222375.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCourtney, J. B. et al. Revisions to the Australian tropical cyclone best track database. Journal of Southern Hemisphere Earth Systems Science, \u003cb\u003e71\u003c/b\u003e, 203\u0026ndash;227. (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChand, S. S. et al. Review of tropical cyclones in the Australian region: Climatology, variability, predictability, and trends. WIREs Climate Change. \u003cb\u003e10\u003c/b\u003e, e602. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/wcc.602\u003c/span\u003e\u003cspan address=\"10.1002/wcc.602\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarper, B. A., Stroud, S. A., McCormack, M., \u0026amp; West, S. A review of historical tropical cyclone intensity in northwestern Australia and implications for climate change trends analysis. Australian Meteorological Magazine, \u003cb\u003e57\u003c/b\u003e, 121\u0026ndash;141. (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlotzbach, P. J., Landsea, C. W. Extremely intense hurricanes: Revisiting Webster et al. (2005) after 10 years. \u003cem\u003eJournal of Climate\u003c/em\u003e, \u003cb\u003e28\u003c/b\u003e, 7621\u0026ndash;7629, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JCLI-D-15-0188.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-15-0188.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCourtney, J. B \u0026amp; Burton, A. D. Joint Industry project for objective tropical cyclone reanalysis: final report, retrieved May 2 2021 at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bom.gov.au/cyclone/history/database/OTCR-JIP_FinalReport_V1.3_public.pdf\u003c/span\u003e\u003cspan address=\"http://www.bom.gov.au/cyclone/history/database/OTCR-JIP_FinalReport_V1.3_public.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCourtney, J. B. et al. Towards an objective Tropical Cyclone dataset for the Australian region. Trop. Cyclone Res. \u003cb\u003e9\u003c/b\u003e, 23\u0026ndash;36. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaulsen, B. M. \u0026amp; Schroeder, J. L. An Examination of Tropical and Extratropical Gust Factors and the Associated Wind Speed Histograms. Journal of Applied Meteorology, \u003cb\u003e44\u003c/b\u003e, 270\u0026ndash;280, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JAM2199.1\u003c/span\u003e\u003cspan address=\"10.1175/JAM2199.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarper, B. A, Kepert, J. D., Ginger, J. D. Guidelines for converting between various wind averaging periods in tropical cyclone conditions. \u003cem\u003eWMO\u003c/em\u003e, [available from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.wmo.int/pages/prog/www/tcp/documents/WMO_TD_1555_en.pdf]\u003c/span\u003e\u003cspan address=\"https://www.wmo.int/pages/prog/www/tcp/documents/WMO_TD_1555_en.pdf]\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnapp, K. R., Kruk, M. C., Levinson, D. H., Diamond, H. J. \u0026amp; Neumann, C. J. The international best track archieve for climate stewardship (IBTrACS) unifying tropical cyclone data. Bulletin of the American Meteorological Society, \u003cb\u003e91\u003c/b\u003e, 363\u0026ndash;376. (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu, C. H., Eizenberg, N. Known issues in the BARRA datasets. \u003cem\u003eBureau of Meteorology\u003c/em\u003e. Accessed April 7 2020 at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bom.gov.au/research/projects/reanalysis/Known_issues_in_BARRA_dataset_July_2019.pdf\u003c/span\u003e\u003cspan address=\"http://www.bom.gov.au/research/projects/reanalysis/Known_issues_in_BARRA_dataset_July_2019.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSimmonds, I. \u0026amp; Murray, R. J. Southern extratropical cyclone behavior in ECMWF analyses during the FROST Special Observing Periods. Weather and Forecasting, \u003cb\u003e14\u003c/b\u003e, 878\u0026ndash;891. (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurray, R. J. \u0026amp; Simmonds, I. A numerical scheme for tracking cyclone centres from digital data Part I: development and operation of the scheme. Australian Meteorological Magazine, \u003cb\u003e39\u003c/b\u003e, 155\u0026ndash;166. (1991).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePepler, A. Record lack of cyclones in southern Australia during 2019. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e, \u003cb\u003e47\u003c/b\u003e, e\u003cdiv class=\"ExternalRefDOI\"\u003e2020GL088488\u003c/div\u003e. https://doi.org/10.1029/2020GL088488. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHosking, J. R. M. L-Moments: Analysis and Estimation of Distributions Using Linear Combinations of Order Statistics. Journal of the Royal Statistical Society, \u003cb\u003e52\u003c/b\u003e, 105\u0026ndash;124. (1990).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, C., Wang, X., Khoo, Y. B. Extreme wind gust hazard in Australia and its sensitivity to climate change. Natural Hazards, \u003cb\u003e67\u003c/b\u003e, 549\u0026ndash;567, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11069-013-0582-5\u003c/span\u003e\u003cspan address=\"10.1007/s11069-013-0582-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2013).\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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"tropical cyclones, extratropical cyclones, wind speed return periods, coastal hazard, reanalysis ","lastPublishedDoi":"10.21203/rs.3.rs-1507699/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-1507699/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLikelihood estimates of extreme winds, including those from tropical cyclones (TCs) at certain locations are used to inform wind load standards for structural design. Here, wind speed average recurrence intervals (ARIs) determined from TC climate data dating back to the 1970s in two quantile-quantile adjusted reanalysis datasets (ERA5 and BARRA [1990]), and best-track observations for context, were compared with Standardized ARIs (AS/NZS) across seven tropical and two subtropical Australian inland coastal regions. Inherent differences between the data used to determine the Standard ARIs (large sample size allow for larger extrapolations; GEV function) and TC data ARIs (smaller sample size and less certain data; the more asymptotic Lognormal/Weibull functions are used) led to the use of different extreme value functions.\u003c/p\u003e\u003cp\u003eResults indicated that although these are two distinct ways of determining design wind speeds, when they are considered equivalent, there was a moderate reproduction of the ARI curves with respect to the Standard in both reanalysis datasets, suggesting that similar analyses using climate model products can provide useful information on these types of metrics with some caveats.\u003c/p\u003e\u003cp\u003eTrends in TC wind strength affecting coastal Australia were also analyzed, indicating a potential slight downtrend in tropical West coast TC wind strength and slight uptrends for tropical East coast TC wind strength, subtropical West coast TC wind strength and subtropical East coast ETC wind strength with low-to-medium confidence. Noting considerable uncertainties, and large interannual variability makes it hard to determine any of these results with a high degree of confidence based on current data.\u003c/p\u003e\u003cp\u003eIt is also difficult to attribute any noted changes in extreme wind gust strength in these datasets to anthropogenic climate change at this time given the relatively short time period and limitations of the data quality including that over longer time periods and preceding the era of widespread satellite coverage in the Australian region. It is hoped that climate models can provide a more homogenous base for these types of evaluations and subsequent projections with respect to climate change simulations.\u003c/p\u003e","manuscriptTitle":"Using historical tropical cyclone climate datasets to examine wind speed recurrence for coastal Australia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2022-04-04 14:38:55","doi":"10.21203/rs.3.rs-1507699/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major revision","date":"2022-05-03T14:31:22+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2022-04-18T20:30:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"266262fa-2fc9-4192-ac73-1d9f5b400aee","date":"2022-04-11T10:35:28+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2022-04-11T02:00:25+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2022-04-01T05:20:48+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2022-03-31T09:43:15+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2022-03-31T09:36:16+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2022-03-31T04:09:08+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"14ef1d83-44fe-4c56-9a76-0b9ba6a27547","owner":[],"postedDate":"April 4th, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2022-06-13T16:59:14+00:00","versionOfRecord":[],"versionCreatedAt":"2022-04-04 14:38:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-1507699","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-1507699","identity":"rs-1507699","version":["v1"]},"buildId":"WrCJVZZCHTDjtuVLN7oU0","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}