What Australian voters didn't know : less extreme rainfall causing floodsFlooding rainy days in decline
On 3 May 2022, the Climate Council published Uninsurable Nation: Australia's most climate-vulnerable places, a report that attracted widespread national media coverage for its claim about homes becoming uninsurable due to natural disasters caused by climate change, primarily riverine flooding.
The report names 27 locations that are the most vulnerable in Australia to an increasing frequency and volume of flood events. Those locations are: Adelaide Ballina Beaudesert Benalla Boonah Brisbane Broadbeach Waters Byron Bay Casino Charleville Cunnamulla Dalby Grafton Helidon Kingaroy Kyogle Lismore Mullumbimby Roma Shepparton St George Stanthorpe Tweed Heads Wangaratta Warwick Winton Yamba (click or tap to a detailed extreme rainfall history page for each individual weather station) Analysis of Bureau of Meteorology rainfall observations at the 27 locations shows a decrease in both the frequency and millimetre volume of the 90th percentile (top 10%), 95th percentile (top 5%) and 99th percentile (top 1%) of all rainfall days when comparing 1915-1999 with 2000-2021. The only climate contributor to flooding is rainfall in high quantities over short durations. The 90th, 95th and 99th percentiles reveal days when extreme rainfall might cause flooding, and this analysis measures their frequency and volume throughout the entirety of each location's weather station observations. Reduced extreme daily rainfall frequency and mm volume is also apparent when comparing 16 of the 27 stations with adequately long records from 1888-1999 to 2000-2021, 22 of the 27 stations from 1899-1999 to 2000-2021, and particularly all 27 stations comparing 1961-1990 with 2000-2021 (1961-1990 is the baseline anomaly reference timeframe for 104 stations that average Australia's temperature trends since 1910 within the Australian Climate Observation Reference Network). If flooding is caused by extreme rainfall, the evidence suggests a decrease rather than increase in the frequency and mm volume of such wet days over the past few decades, raising questions about the validity of claims that climate change is causing an increase in the risk of flooding at what are said to be the 27 most vulnerable locations that have long-term rainfall observations. Extreme rainfall decline Inaccurate forecasts of increased flood risk as claimed by the Climate Council and supported by the insurance industry may result in unjustified influences on development planning policies, the real estate value of homes owned by hundreds of thousands of Australians, and the lifestyle location choice of people wishing to purchase property. Voter memories of the February and March floods in NSW/Queensland (see NSW flood inquiry submission and Institute of Public Affairs report showing decreased extreme daily rainfall in affected region) were a strong influence on the so-called climate change election of 2022 which resulted in a surge in Green votes and an unprecedented number of seats won by independent teal candidates with similar climate change policies. Cost of living was a key election issue but surveys showed up to 40% of voters in some electorates had climate change as a priority, and both their primary and preference votes favoured Greens and teals across Australia. Voter attitudes and fears were reinforced by widespread national media coverage of the Climate Council claim on 3 May that one in 25 Australian homes will be uninsurable by 2030 mostly because of increased flooding caused by climate change. The influence of the Climate Council's flood insurance claims have received global attention, as witnessed in a BBC report headlined Australia election: How climate is making Australia more unliveable. What voters didn't know Four days after the election, the Bureau of Meteorology published Special Climate Statement 76 – Extreme rainfall and flooding in south-eastern Queensland and eastern New South Wales which mostly confirms the floods were not due to climate change but nevertheless states: Observations show that there has been an increase in the intensity of heavy rainfall events in Australia. The intensity of short-duration (hourly) extreme rainfall events has increased by around 10 per cent or more in some regions in recent decades, with larger increases typically observed in the north of the country. It is doubtful the bureau has hourly rainfall records from its weather stations dating back more than a few decades, and hourly rainfall is responsible for localised flash flooding rather than broad regional flooding as caused by daily rainfall. It is noteworthy that the bureau cites unknown hourly recordings not readily accessible to the public, rather than daily recordings as used in this analysis. The Climate Council agrees that Australia's 2022 federal election was strongly influenced by the NSW/Queensland floods two to three months earlier ... see THE 2022 CLIMATE ELECTION: Unpacking how climate concerned Australians voted. The 2022 floods were devasting and in some locations record-breaking, but long-term data analysis of rainfall observations in the 27 nominated cities and towns shows that both the frequency and volume of extreme daily rainfall that causes flooding have had a decreasing rather than increasing trend in recent decades. The 2022 floods were more likely caused by saturated soils following two consecutive La Ninas and land clearing increasing rainwater runoff, with some influence from hard surface infrastructure that channels floodwaters and prevents soil absorption. A prime driver of Australia's 2022 federal election result was voter fear that climate change is happening in their backyard and causing more natural disasters such as floods, but rainfall records show that fear is unwarranted and misplaced. Summary table This page presents analysis of the 27 weather stations in the Climate Council's nominated vulnerable towns and cities, with their annualised daily frequency, millimetre totals and averages within four timeframes for 90th, 95th and 99th percentile rainfall up to 2021, as well as spreadsheets providing more detailed information. The table below summarises the research findings for 90th, 95th and 99th percentile rainfall days when comparing 16 of the 27 stations with adequately long records from 1888-1999 to 2000-2021, 22 of the 27 stations from 1899-1999 to 2000-2021, all 27 stations from 1915-1999 to 2000-2021, and all 27 stations from 1961-1990 to 2000-2021.
The table shows a consistent increase in the annual average frequency of all rainfall days (from 0.1mm to the wettest), a decrease in the average annual mm volume of those days, and a decrease in the annual average frequency and total mm volume of wet, very wet and extremely wet days within the 90th, 95th and 99th percentiles. Quick link menu for charts below Wet 90th percentile rainfall days (top 10%) 90th percentile rainfall at 16 stations since 1888 90th percentile rainfall at 22 stations since 1899 90th percentile rainfall at 27 stations since 1915 90th percentile rainfall at 27 stations since 1961 Very wet 95th percentile rainfall days (top 5%) 95th percentile rainfall at 16 stations since 1888 95th percentile rainfall at 22 stations since 1899 95th percentile rainfall at 27 stations since 1915 95th percentile rainfall at 27 stations since 1961 Extremely wet 99th percentile rainfall days (top 1%) 99th percentile rainfall at 16 stations since 1888 99th percentile rainfall at 22 stations since 1899 99th percentile rainfall at 27 stations since 1915 99th percentile rainfall at 27 stations since 1961 90th percentile rainfall days 90th percentile rainfall days at 16 locations comparing 1888-1999 with 2000-2021
Return to top of page or chart menu 90th percentile rainfall days at 22 locations comparing 1899-1999 with 2000-2021
Return to top of page or chart menu 90th percentile rainfall days at 27 locations comparing 1915-1999 with 2000-2021
Return to top of page or chart menu 90th percentile rainfall days at 27 locations comparing 1961-1990 with 2000-2021
Download 90th percentile Excel spreadsheet for 27 weather stations Return to top of page or chart menu 95th percentile rainfall days 95th percentile rainfall days at 16 locations comparing 1888-1999 with 2000-2021
Return to top of page or chart menu 95th percentile rainfall days at 22 locations comparing 1899-1999 with 2000-2021
Return to top of page or chart menu 95th percentile rainfall days at 27 locations comparing 1915-1999 with 2000-2021
Return to top of page or chart menu 95th percentile rainfall days at 27 locations comparing 1961-1990 with 2000-2021
Download 95th percentile Excel spreadsheet for 27 weather stations Return to top of page or chart menu 99th percentile rainfall days 99th percentile rainfall days at 16 locations comparing 1888-1999 with 2000-2021
Return to top of page or chart menu 99th percentile rainfall days at 22 locations comparing 1899-1999 with 2000-2021
Return to top of page or chart menu 99th percentile rainfall days at 27 locations comparing 1915-1999 with 2000-2021
Return to top of page or chart menu 99th percentile rainfall days at 27 locations comparing 1961-1990 with 2000-2021
Note that all chart trendlines are polynomial 4. Usually if the third chart of average mm shows an increase, it is because of a disproportionate reduction in frequency and volume of extreme rainfall days ... i.e the proportionate reduction in the average number of extremely wet days was considerably greater than the average reduction in their mm volume of rainfall (e.g. 10/2 = 5 but 9/1.7 = 5.3). Both the frequency and volume of extreme rainfall days can decline but if the frequency decline is greater than the volume decline, it results in a higher average. Download 99th percentile Excel spreadsheet for 27 weather stations Return to top of page or chart menu Addendum 1 The results above can be compared with BoM data showing the annual fractions (percentages) experiencing historically high (decile 10 or top 10%) rainfall since 1900 across Australia, Eastern Australia and Western Australia.
The BoM data shows a decline in decile 10 downpours since the 1970s, with Western Australia an exception albeit with a decline since the 1990s. Alternatively and in light of the February and March 2022 floods in NSW and Queensland, BoM data shows the annual fractions (percentages) experiencing historically high (decile 10 or top 10%) rainfall since 1900 in those states.
Download BoM 10th decile Excel calculations The percentage area of NSW and Queensland experiencing decile 10 rainfall each year has declined in recent decades, with February/March decile 10 rainfall increasing but still below the 1970s average and Feb/Mar 2022 seeing 8.30% of area affected which is below the 1900-2022 long-term average of 10.57%. Addendum 2 Lismore in New South Wales was heavily impacted by the 2022 floods but it is well established that the city has largely been built on the floodplains of two intersecting rivers and has a history of flooding. Although flooding is mostly caused by rainfall upstream of Lismore, it is nevertheless worth considering the city's previous rainfall records. Lismore Airport's highest daily rainfall in February or March 2022 was 146.8mm. Lismore's flood level peak was 14.37 metres in February 2022, a record. From 1885 to 2021, Lismore's highest 99th percentile rainfall day was 209.8mm in 1989. Below charts 30 year climate period data since 1884 extracted from the BoM's Monthly Climate Statistics for Lismore (Centre St) and Lismore Airport AWS, averaged for seasonal comparisons and showing a decrease rather than increase in the mm volume of such days with the potential to cause flooding.
Addendum 3 In November 2022, the journal Science published Intensification of subhourly heavy rainfall (paywalled but authors' explanatory article at The Conversation) which found an increase of at least 40% in the rate at which rain falls in the most intense rapid bursts (defined as very heavy rainfall lasting less than an hour and as brief as 10 minutes) in Sydney from 1997 to 2018, based on weather radar observations at Newcastle, Terrey Hills and Wollongong. The study found rainfall rates per decade for the top 5% of rapid rain bursts increased 20.07% at Wollongong, 49.18% at Newcastle and 50.84% at Terrey Hills. From the study's supplementary notes ... All radars have experienced hardware and configuration changes in their operational periods. The most important one is a configuration change in all radars from a 10-minute volume cycle to a 6-minute volume cycle, which occurred on different dates (December 2012, June 2009 and June 2013, respectively for the three radars). The study attempts to correct for these instrument changes that caused an increased radar resolution able to better detect rapid rainfall bursts but, coupled with the brevity of a 1997-2018 climate period with no historic preceding observations, there are valid questions about the consistency and reliability of its claim that Sydney has witnessed an alarming increase in the risk of flash flooding due to climate change. The authors' article states: Despite the dramatic increase in the intensity of rapid rain bursts, we found little if any change in daily or even hourly rainfall over the same region. Terrey Hills is an automatic weather station with no historic daily rainfall observations prior to its start year of 2004, so the charts below substitute Terry Hills with Sydney (Observatory Hill) itself in combination with Newcastle and Wollongong to show the frequency and total/average mm volume of 90th (wettest 10%), 95th (wettest 5%) and 99th (wettest 1%) percentile rainfall days from 1895 to 2021: 90th percentile rainfall days at Newcastle, Sydney and Wollongong comparing 1895-1999 with 2000-2021
95th percentile rainfall days at Newcastle, Sydney and Wollongong comparing 1895-1999 with 2000-2021
99th percentile rainfall days at Newcastle, Sydney and Wollongong comparing 1895-1999 with 2000-2021
The historic data shows a 16.7% decrease in 90th percentile frequency, a 20.7% decrease in 95% percentile frequency, and a 21.4% decrease in 99th percentile frequency of extreme rainfall days that might cause riverine flooding in the Sydney region when comparing 1895-1999 with 2000-2021, as well as a consistent reduction in the average mm volume of such events and an average reduction in their annual total mm volume. Despite questions about the reliability of the study's data source, it may be that Sydney has seen an increase since 1997 in extreme rainfall lasting up to an hour that can cause localised flash flooding. However, reliable daily rainfall observations since the 1800s show that Australia's largest city is enjoying a reduced risk of major flooding despite a popular public belief that climate change is causing an increased risk. Download Excel spreadsheet with daily rainfall data for Newcastle, Sydney and Wollongong (includes Terrey Hills although not included in the averages presented above)
This website produced by Scribeworks 2008-2023 |
