Professor David has propounded an interesting theory about prehistoric climatic changes in Australia. [He] inclines to the theory of Dr. Arrhenius , that changes of climate which control the rainfall depend largely on the variation of the amount of carbon dioxide present in the earth's atmosphere. This gas acts as a most efficient blanket for the retention of the earth's heat, received chiefly from the sun. If the present amount of carbon-dioxide were doubled a mild climate would prevail even at the north and south poles, and the polar ice caps would probably disappear, and the temperature in other parts of the world would be raised. If, on the other hand, the quantity were reduced by a trifle less than half its present amount, conditions like those of the last great ice age would return.
Wanganui Chronicle, 19 June 1906.
Climate change is like the growing stench of a dead possum under the floorboards. But it’s a very old possum, first suspected of being there in the late-1800s. However, it wasn’t until the mid-1900s that the first whiffs wafted up. And it was only in the late-1970s and early 1980s that a strong consensus emerged amongst climate scientists that the smells were real and were going to get rather pungent in the relatively near future. Others, led by the fossil-fuel industries, told us we were imagining things - or that it wasn’t so bad, and that we just needed stronger air fresheners. But by the late 1900s the stench was unmistakable. Many people alive today have lived with it for all their lives and have decided that something really does need to be done.
Earth’s average surface air temperature depends on the solar radiation we receive at the top of the atmosphere, how much of that radiation is reflected back to space, and the characteristics of the “thermal blanket” that the atmosphere provides. The latter in turn depends on the concentrations of the greenhouse gases. These are largely transparent to solar radiation but are collectively very good at absorbing almost all the longwave radiation emitted by earth’s surface – and reradiating much of the energy back to the surface.
Human activity, especially since the industrial revolution, has increased the concentration of several greenhouse gases, especially carbon dioxide (CO2). This has enhanced the cycle of longwave radiation, leading to a warming of the lower atmosphere. As we progress through this century Earth is expected to continue to warm, but the magnitude of that warming is highly uncertain, mainly because we don’t know how concentrations of greenhouse gases will evolve, and because of uncertainty related to climate sensitivity to those increases. The former depends on factors affecting emissions, such as population and economic growth, and increasingly on political decisions (or lack of them) related to reducing emissions.
Because we cannot know what future emissions will be, the approach taken is to derive plausible scenarios based on a range of mutually consistent assumptions. At one extreme we could consider a world of high population growth, high economic activity, and little progress in terms of mitigation (e.g. decarbonisation). At the other, we can imagine a future of heightened political action, leading to net-zero emission by the middle of this century. You may think that neither of these scenarios is truly plausible, but they do usefully bracket the range of what the future might hold, and mid-range scenarios can be derived to investigate other possibilities.
The figure below shows IPCC global warming projections through to the end of this century. Lines are best estimates, and their divergence indicates uncertainty related to progress (or not) towards net-zero emissions. Vertical bars show additional scientific uncertainty ranges for the two extreme emissions scenarios, mainly related to that thorny question of climate sensitivity. The grey shading shows the full uncertainty range.
Global warming projections for this century (relative to 1995–2014) for the five IPCC core emissions scenarios. Large circles are best estimates for the 20 years straddling the year shown. Vertical bars show uncertainty ranges for the “Very high” and “Very low” scenarios. Small crosses and dots (bottom-left) show observed annual global temperatures and 20-year averages, also relative to 1995–2014.
Data: Lee et al (2021) - IPCC AR6, Chapter 4, Table 4.5 (p. 582) and NASA (global temperature observations).
Notes. The overlapping uncertainties means that global warming of 1.1°C by 2050 could plausibly be associated with any of the five emissions scenarios. It takes several decades for the lines to diverge because warming in the immediate future will be dominated by greenhouse gases that are already in the atmosphere.
The 2015 Paris Agreement set a target of keeping global warming to well below 2°C and preferably less than 1.5°C of pre-industrial temperature. In this context, only the two low-end scenarios (net-zero emissions by mid-century) would appear to have any hope of meeting either target. Moreover, the Low scenario would require climate sensitivity to come in at the lower end of its uncertainty range.
The first regional climate change scenarios for New Zealand were produced in the late 1980s and have been updated, mainly by NIWA climate scientists, through the six IPCC assessment cycles. The following is based on the results of NIWA’s regional climate modelling downscaling, which started to be made publically available mid-2024.
The figure below shows shows projections of Auckland City warming for this century (the three solid lines). For comparison, the figure also includes the global-scale projections for the three scenarios (dotted lines) and the uncertainty range (shaded area) across the five core IPCC scenarios (from the figure above). The similarities suggests that the full global range is probably a reasonable estimate for local Auckland warming by late century: less than 0.5°C through to about 4.5°C.
Auckland warming projections for this century (solid lines) compared to global projections and uncertainties (dotted lines and shaded area, both from the previous figure). Large circles are best estimates for the 20 years straddling the year shown. Small crosses and dots show observed annual temperatures and 20-year averages. All data are anomalies (differences) relative to 1995–2014.
Data: NIWA dynamical downscaling (Auckland scenarios), Lee et al (2021) – IPCC AR6 (global scenarios), National Climate Database (observations).
Average temperature is all very well, but it isn’t something we experience. Much more interesting is how a warming world will affect temperatures at the start of the day and mid-afternoon, especially in summer and winter. And most of us don’t live in the city centre. A closer look at the downscaling results suggests that:
• Summer and autumn will warm more than winter and spring.
• Days will warm more than nights.
• There are a few wrinkles related to topography and proximity to the coast.
• For daily maximum temperatures, most of the region will warm slightly more than the city centre, especially in summer.
Contrary to these modelling results, the observational data shows us that nights have warmed more than days (there in a relevant figure on the Present page).
The map in the figure below shows late-century projected change in annual rainfall for the High emissions scenario. Everywhere is projected to dry by at least a few percent, with up to 10% less rainfall in the north and over the Hunua Ranges in the southeast.
Late-century projected changes in rainfall (%). The map shows annual rainfall change for the High emissions scenario. The plots on the right show projected changes for the city and the Hunua Ranges across all scenarios and seasons, for three 20-year time periods centred on the dates shown.
Data: NIWA dynamical downscaling.
Notes. Each bar in the right panels corresponds to a season, or the full year, and covers the spread of results for the three scenarios. The small white dots show results for the Mid-range scenario. The three sets of bars for each location are for 20-year blocks centred on the year shown. The map plots values for all grid points corresponding to the most negative end of the solid black (Annual) bars for 2090 (2081–2099 average) – i.e. High scenario results.
More detailed results for the city centre and for the Hunua ranges are shown in the right-hand plots in the above figure. In summary:
• Changes in the seasonal pattern of rainfall are similar across the region.
• Uncertainty increases through the century.
• The direction of change is uncertain for most seasons, except for spring from mid-century.
• The evidence for drying increases through the century. Note how the centre of gravity of the bars shifts to below the zero-change line.
• Spring changes are the largest and have the strongest drying signal.
• Annual drying is dominated by that strong spring signal.
Other downscaling results indicate that the projected drying trend is caused by an increase in the number of dry days (<1 mm).
Global warming is also expected to result in more extreme storms. Projected changes are unequivocal in terms of direction and are potentially very large. The changes are greatest for short duration intense bursts of rainfall.
Future climate change will affect us in multiple and highly diverse ways. Somewhat perversely, as a nation dependent on international trade, some of the most important impacts are likely to relate to how climate change affects our trading partners, rather than what happens locally. It will be small comfort if New Zealand warms less rapidly than most of the northern hemisphere if impacts there see significant impacts on our markets. We may also see emergent issues related to climate refugees, caused by famine and rising sea levels, and possibly water wars. Sea level rise is an inevitable national issue, including that vexing question: who pays?
The above issues, and many others like them, will keep politicians, policy makers, and climate change activists (and denialists) busy in the decades ahead. But the focus here is Auckland, so this section deals with how regional-scale climate change will impact us locally. An example related to potential impacts on heat stress is given below (the book has a couple more: household water tanks and the urban water supply. Some other potential impacts are then briefly outlined.
Temperature extremes in Auckland are relatively rare. We do occasionally exceed 30°C, but this pales into insignificance compared to maximum temperatures routinely reported elsewhere. For example, in June 2024 alone, temperatures in the mid-to-high forties, and even above 50°C, were recorded in multiple Northern Hemisphere locations, in some cases associated with hundreds of deaths.
Heat stress becomes an issue in Auckland when summer temperatures in the mid-to-high twenties coincide with high relative humidity, typically associated with northerly winds. We suffer under these circumstances because higher relative humidity reduces the efficiency of our bodies primary cooling mechanism: sweating. Because the two variables are inter-related, the combined effects of temperature and relative humidity are commonly presented using a heat index, where the influence of relative humidity is converted to a temperature equivalent. Using this approach, 29°C at 40% relative humidity has a similar heat index to 26°C at 100%. The higher relative humidity of the cooler case makes it seem about as hot as the warmer one because water pools on our bodies, but essentially no evaporation occurs. The physiological stress our bodies feel will be similar for the two cases.
The figure below is an example of a heat index. Temperature and relative humidity are the two axes of the graph, and the heat index is represented by the colours in the body (scale bar on the right). The black stepped lines across the plot break it into five warning zones with associated health impacts listed below the graph. Note how the vertical (temperature) spacing of the lines reduces towards the lower-right corner of the plot. This tells us that heat stress is more affected by an increase in temperature when relative humidity is higher.
Temperature and relative humidity converted to a heat index – the latter represented by the colours in the body of the graph (scale to right). Stepped black lines divide the plot into the five warning zones described in the bottom box. Hourly data (24°C ≥ Tmax) from the MOTAT climate station are superimposed (March 2016 – February 2024). Circles highlight 12 February 2022, as recorded (with dots) and with 4°C added (no dots).
Data: Lu & Romps (2022) Extended Heat Index (°C values from authors’ online calculator), National Climate Database (observations).
The heat index plot and the associated warning zones apply to a very specific demographic: well hydrated, young, fit adults, who move into the shade when it gets hot, and strip down to their birthday suits if needs be. For the poorly hydrated/unfit/elderly/modest the health impacts are worse, and very much worse if several of these apply. The very young and the elderly are most at risk.
The black dots superimposed on the heat index plot are hourly maximum temperature and relative humidity data from NIWA’s MOTAT climate station (4 km southwest of the city centre) for eight non-calendar years (March 2016 to February 2024). Hours with temperatures less than 24°C are not plotted (all Safe) and many dots represent multiple hours. Circled dots in the lower-right corner denote 12 February 2022, a high humidity day with temperatures in the mid to high twenties (some hours overlay). Circles without dots show these same data points translated by an arbitrary +4°C.
Overlaying the MOTAT data onto the heat index plot provides a current climate baseline for heat stress for those young, fit and immodest adults. For the eight years analysed, about 141 hours per year were in the Caution zone, with <1 per year in the Extreme Caution zone. Most occurred early in the afternoon of the four warmest months (December to March).
The figure below shows the sensitivity of human heat stress in Auckland to increased hourly maximum temperature. Caution zone hours more than double with 1°C warming and increase more than eight-fold at +4°C. Extreme Caution hours appear in low numbers with 1°C warming and increase significantly with further temperature increases. Danger hours first occur at +3°C warming, rising to about three per year at +4°C, which is also when Extreme Danger hours appear. The latter all occur in mid-February and have relative humidities of 84–94%.
Most Aucklanders are probably uncomfortable when the heat index is in the Caution zone. This sensitivity analysis suggests that as we move towards another 1°C warming these unpleasant conditions will occur twice as often – and many of those future hours will be much more uncomfortable than we are used to. For those who live to see 2°C warming, a further doubling is indicated. The changes are large enough to be concerning and are likely to be serious for the young/unfit/elderly. Determining how serious requires a more sophisticated approach in terms of how future climate will evolve and medical expertise related to the physiology of most vulnerable. Afterall, most of us are not young, fit, and immodest!
Impact of increasing temperature on heat-stress hours (top four heat index safety zones). Zero on the bottom axis represents the baseline period (March 2016 – February 2024). The inset highlights Danger and Extreme Danger counts for 3°C < Tmax (dotted box). Note that the left axis of the inset has a log scale.
So, what does the future have in store for us? The (sort of) good news is that projected changes are modest through to mid-century. Even for the High scenario, projections are for less than a doubling of heat stress hours, with none in the Danger or Extreme Danger zones. If global emissions follow the Low emissions scenario and climate sensitivity does not turn out to be on the high side, then late-century heat-stress days may also be similar to recent experience. However, the situation will be much more serious if the net-zero emissions track is not achieved by mid-century. For the High scenario, a five-fold increase in heat-stress hours is plausible, including an increase in Extreme Caution hours from an average 0.4 to 34 hours/year. If climate sensitivity comes in at the high-end of estimates, a nine-fold increase in heat-stress hours is plausible, including an average 95 Extreme Caution hours/year and some Danger hours (1.5/year). If you have lived in Auckland for a while, you can get a feel for the scale of the projected changes by recalling the worst heat-stress day you have experienced (perhaps 12 February 2022). Now imagine weeks of that!
The list of potential climate change impacts is long and is across multiple facets of the human and natural environments. They range from near certain to highly speculative. Some we can be confident about, but the details may elude us. All are uncertain in terms of scale because we don’t know how quickly we will achieve net-zero emissions (if at all) and because significant scientific uncertainty remains, especially that sticky question of climate sensitivity. Here’s my short list of potential impacts that we can expect, and should be planning for:
• Sea level. Auckland Sea level has increased by about 0.2 m since the beginning of last century and is projected to increase by at least another 0.4 m by the end of this. Related impacts include increased coastal erosion and inundation.
• Health. Warmer winters will ease respiratory issues associated with cold and damp houses. But subtropical pests and diseases will be more prevalent or become established for the first time.
• Energy use. The timing of peak energy use will shift (less winter heating, more summer cooling).
• Extreme rainfall. We can expect more frequent and intense flash flooding and landslips, and to pay for major remedial work on stormwater drains (or alternatives).
• Drought. The causes of droughts will be influenced by climate change but how this will manifest in terms of drought frequency and intensity is uncertain.
• Agriculture. Livestock will be detrimentally affected by heat stress and will benefit from fewer cold extremes. Regional drying will reduce pasture productivity and increase irrigation needs. There will be new opportunities in crop production, but some crops may cease to be viable.
• Fire. Summer drying will increase fire risk.
• Natural environment. Ecosystems will be significantly affected and will evolve.