Present

Early Climate Observations

Europeans brought many things to Aotearoa, including a new name, trade, alcohol, muskets, writing, knowledge of the outside world, Christianity, new biota, and devastating diseases. They also brought the first scientific instruments to record weather. By the mid-1800s several meteorological observatories had been established around New Zealand, including one in central Auckland. The era of systematic daily observations had begun.

Central Auckland has one of New Zealand’s longest continuous records of daily meteorological observations. There are some earlier fragmentary records, but a useful continuous record, at least for rainfall, dates to observations by the Royal Engineers in Albert Barracks, starting in 1853. Observations continued there until 1868, when the observatory was relocated to Auckland Domain under the auspices of the fledgling New Zealand Meteorological Service (NZMS). It was moved again in 1883 to the roof of the old Princes Street Museum.

Finally, in 1909 the observatory was moved to a circular enclosure in Albert Park, very close to the original Albert Barracks location. NZMS closed the Albert Park station in 1994, again because of vandalism, but it has continued as a rainfall site, currently under the auspices of Auckland Council.

Additional Auckland meteorological observatories were commissioned in the 1900s, driven by the requirements of aviation – initially flying boats on the Waitemata and then the major airports at Whenuapai, Ardmore, and Mangere. Others were added by Crown research organisations. The number of meteorological variables increased, as did the sophistication of the instruments used, and from the late 1900s there was a major shift from manual to automatic instrumentation. Also noteworthy was the development of a much denser rain gauge network by Auckland’s various iterations of local government and agencies. However, for studies related to past climate change, it is the long and relatively basic records that are most relevant (and the focus here).

Temperature

Auckland has surface air temperature data dating back to 1853. The early part of that record in Albert Barracks is of uncertain quality because we don’t know how the thermometers were screened, and some details of the site environment are uncertain. There is also some uncertainty about possible screen changes prior to 1950. However, from May 1868, we do appear to have a fairly good quality record, albeit from multiple sites which need to be stitched together. The relevant needlework was first undertaken by Jim Salinger over 40 years ago as part of his PhD research and is now routinely done by NIWA as part of their New Zealand 7-station series.

The graph below shows annual average Auckland temperature since 1869. It is a composite of the sites listed across the top, with Albert Park split in two by a thermometer screen change at the end of 1950. Note that the so-called reference site is Auckland Airport, with all other data adjusted to it – so in essence the time series is an estimate of what temperature at Auckland Airport would have been if recordings had been taken continuously there from 1868.  However, there is one important caveat here because the site adjustments do not account for urban warming, estimated by Brett Mullan, who led the NIWA work, to be about 0.3°C.

It appears that Auckland’s average temperature declined through the last decades of the 1800s to about 14.2°C near the beginning of the 1900s. It then increased, with a few flat spots, through to the present. The first 1°C increase was reached by 1980 (included most of the 0.3°C related to urban warming), with roughly a further 0.6°C since.  Variations of average annual temperatures about the smooth curve in the graph (about ±1°C) are mostly caused by changes in the frequency of north/south wind flow.

The time series of average annual temperature is a useful starting point, but the instrumental record is richer than this and a little more analysis can provide answers to questions such as: are summers warming faster than winters or is it the other way around; is it warming more during the day or at night; and how significant are these changes anyway? Daily average temperature is calculated by averaging the daytime high and the night-time low. A simple analysis of these data answers all three of the questions posed above.

The graph below compares monthly maximum and minimum temperatures for two 30-year blocks near the start and end of the available data. Note how the lines for the early period (dotted) always plot below those for the later one (solid). This tells us that temperatures have increased in every month for both day and night. Nights have warmed about 0.5°C more than days, although how much so varies through the year. Also, summers have warmed a bit more than the other seasons.

Rainfall

Daily rainfall was recorded in Auckland prior to 1853, but data are presently only available from when the Royal Engineers began observations in Albert Barracks in that year. Some of the daily data have been lost, but all monthly totals are available, albeit from multiple sites. Recorded rainfall is highly sensitive to site exposure, with more exposed rain gauges systematically recording less. Because of this, the multiple site changes introduced major inhomogeneities into the long-term record which must be adjusted for before it can meaningfully be used to investigate changes in Auckland’s rainfall regime over the last 170 years. To this end, I constructed two central Auckland rainfall series between 2020 and 2023, first in response to the COVID-19 drought (monthly) and then the early-2023 deluge (daily). These are the two data sets used in this section (Fowler 2020, 2021a, 2021b, 2023a, 2023b).

Drought

The graph below is a 12-month running sum through the monthly rainfall series, showing the history of dry and wet periods for the last 170 years. Droughts are the large and prolonged deviations of the series below the long-term average. They are highly variable in both intensity and duration, but a few features are worth picking out. About once in 10 years, 12-month rainfall has declined to 30% below average (but never 50%). Some of those droughts are single-year events while others linger for several years, occasionally with multiple peaks. There also appears to be a trend to more frequent droughts, and fewer wet episodes, starting in the 1970s. Finally, some droughts are broken abruptly – indeed by deluges in the case of the World War 1 (WW1) and COVID-19 droughts.

The COVID-19 drought was one of the most significant in Auckland’s history. It was comparable in size to the WW1 drought which had provided a useful benchmark for the incremental development of the city’s water supply reservoirs in the Waitakere and Hunua ranges through the 1900s. It was the culmination of several dry decades, killed off numerous trees, and caused enormous stress for many people relying on household water tanks. The water reservoirs dropped to levels not seen for a generation, Watercare dusted off the drought management plan, and we came out of lockdown being urged to conserve water (but to keep washing our hands). The drought also engendered yet another political panic about Auckland’s water supply system. Mayor Goff declared a “water crisis” and we rapidly transitioned from “keep calm and carry on” to “the sky is falling – who is to blame”. The sky didn’t fall, but heads did roll, before the drought broke.

Deluge

From drought in mid-2021, Auckland’s climate reversed course to extreme wetness by early 2023. Such drought-to-deluge flips are a feature of Auckland’s climate and even the scale of this latest one wasn’t unprecedented. But in early 2023, Aucklander’s had no collective memory of such events, as a miserable start to summer turned into the shock of the monumental 27 January storm on Auckland Anniversary Day. So, what exactly happened and what can the historical record tell us about how rare these events are?

Autumn 2022 was dry with 12-month rainfall still 12% below average, at the tail end of the COVID-19 drought. But winter was wet, bringing cumulative rainfall well above average and likely replenishing any residual soil water deficits. Spring was also wet and what seemed like endless rain continued through December and early January. Aucklanders were moaning about a missing summer. There was a week-long respite leading up to Auckland Anniversary Day which raised hopes – but then the rain returned.

Rain started falling in the city in the early hours of 27 January and continued at moderate intensities for 13 hours (below). By 4pm, 81 mm had been recorded and if it had stopped raining then we would have been talking about a notable event with a return period of about five years. But it didn’t stop. Three hours of torrential rain added another 181 mm and by the end of the day a new daily rainfall record had been set (281 mm). The extreme rainfall intensities during the early evening caused extensive flash flooding across the Auckland region. Several lives were tragically lost, thousands of properties were severely damaged, and there was widespread damage to land caused by slips.

Given the magnitude of the storm and the rainfall intensities over just a few hours, extensive urban flooding was inevitable. However, the flooding was made much worse by the exceptionally wet leadup to the storm. Soils are normally at their driest at this time of year and have the capacity to absorb much of the rain associated with big summer storms. But the usual summer drying cycle through to late January was very weak and the 81 mm of rainfall to 4pm essentially brought the entire city to saturation. When the most intense rainfall fell in the early evening almost every surface across the city produced runoff. Catastrophic flash flooding was the result.

The Anniversary Day storm was undoubtedly a monster, but just how significant was it? To answer this question hydro-climatologists turn to something called return period analysis. This shows that the storm was a 50+ year event for all rainfall durations greater than 10 minutes and several hundred years for durations covering the most intense 2–3 hours in the early evening. It was certainly the largest daily rainfall ever recorded and was responsible for January 2023 being the wettest month – by a very large margin (below).

The 27 January storm will long be remembered by Aucklanders who experienced it, but it wasn’t the only extreme rainfall event of the start of 2023. For many living on the region’s west coast, the 13 February storm (ex-tropical cyclone Gabrielle) will be a more vivid memory, because of the lives lost and the damage it caused to coastal communities. It is also noteworthy that the first half of 2023 set another record in terms of the frequency of extremes, with four 80+ mm rain days – more than any other full year.

Drought-to-Deluge Flips

The storms in the first half of 2023 broke yet another deluge record (by a whisker) – this time the 12-month rainfall total. The previous record was set in September 1917 at the height of another deluge that followed on the heels of the WW1 drought. Such drought-to-deluge flips appear to be characteristic of Auckland’s climate and overlaying the two events reveals striking similarities. Both deluges were preceded by droughts, each drought took about two years to peak, then a year or so to return to normal. In both cases the flip into deluge required only another six months. Overall, the impression one gets is of similarity rather than difference and a “history repeats” conclusion seems reasonable. 

However, there is one notable difference between the WW1 deluge and our most recent experience: the COVID-19 drought lasted a little longer but then flipped to deluge more quickly. The new record was broken 10 months after the long-term average was exceeded, about twice as fast as was the case in 1917. The chief culprit was the 27 January storm. Although both deluges had two 100+ mm daily rainfall totals, this storm produced 81% more rainfall than the largest WW1 storm and without it we wouldn’t have come close to a new record. Aucklanders would have been moaning about a terrible summer and mopping up, not dealing with the aftermath of catastrophic flash flooding, and worrying that perhaps our climate has flipped to a new normal caused by global warming.

Additional Information

Surface air temperature: what it is, why the average isn’t, and why that’s OK

The curious case of Albert Park’s 1950 screen change

The NIWA 7-station temperature series