Enhanced Supercomputing Capabilities Increase Confidence in Latest Round of IPCC Climate Projections
What do the latest IPCC climate projections mean for Ireland?
The latest climate assessment report from the Intergovernmental Panel on Climate Change (IPCC) is the 6th in a series, going back as far as 1990. The messaging from this latest report couldn’t be clearer, with UN secretary general António Guterres calling it ‘A code red for humanity’. This is a long way from the more nuanced and tentative message from the first such report in 1990 that human activity and emissions of greenhouse gases will result in a likely increase in global-mean temperature.
The IPCC keeps asking the question, what will the future climate be like under various different emission scenarios? Ever since 1990, scientists have been giving the same answer: the more greenhouse gases we emit, the warmer the world will get. The only difference over the past 30 years is that as the data builds up, as the science progresses, as the models get better and computers get more powerful, there is more and more confidence in that answer. What was “likely” just a few years ago is now “virtually certain”.
Advances in both climate science and supercomputing have enabled the continuous improvement of climate models. The 1990 IPCC First Assessment Report assessed small ensembles of simple atmosphere-only climate models with typical grid resolutions of 500km. In contrast, the IPCC AR6 reports assesses very large ensembles of high-resolution (~80km) climate models which include multiple coupled components of the climate systems such as the atmosphere, ocean, sea ice, land surface, dynamic vegetation, atmospheric chemistry and ocean biochemistry. These improvements have resulted in more detailed and accurate representation of the climate system, and allowed the IPCC to assign higher confidence levels to climate projections
Ireland’s input to IPCC climate projections
One way that Ireland contributes to the IPCC reports is through its participation in a consortium of European countries that develop and run one of the 30 or so global climate models that feed into the IPCC assessments.
As part of this consortium, Dr. Paul Nolan of the Irish Center for High-End Computing (ICHEC) has conducted many centuries worth of global climate simulations with the ‘EC-Earth’ climate model. As an “earth system model”, it represents all of the most relevant physical processes that operate within and between the atmosphere, the ocean, the land surface, and sea ice. Simulations range from the historical past, so that model performance can be compared with real climate records, out to the end of the 21st century, to provide a picture of what future climates could be like. The research is funded and supported by the EPA, Met Éireann and the Marine Institute and is carried out in collaboration with Met Éireann.
How is it possible to simulate the climate in 2050, 2100, or any other future year?
Starting from rest, or perhaps some other state that approximates current conditions, the model earth is set spinning, and heating from the sun is turned on. During a long spin-up phase, the forcing and boundary conditions are held constant (apart from the daily and annual cycles) at values representative of pre-industrial times (1850, say). Over many simulated years of such regular forcing, the model atmospheric and oceanic circulations evolve into a more or less equilibrium climate with no long-term trends. (The deep bulk of the ocean is often an exception: this has such massive heat capacity and such slow circulations that it takes centuries to equilibrate).
Once the model has settled into a long-term climate balance, the historical clock can be started, and forcing appropriate to each historical year can be applied, from 1850 right up to the present day. The governing equations are integrated numerically, which allows the model state to advance through time in discrete incremental steps of just 30 minutes or so. In practise, the model state is saved at regular intervals, and these saved “snapshots” can be used to restart the model to run for another few years.
Changes in climate forcing (primarily through changes in greenhouse gas and aerosol concentrations) are relatively well known over the past couple of centuries, and particularly well known since 1950 or so. We can think of the climate of 1990-2020 (whether real or simulated) as the response to the climate forcing up to the end of that period. What future climate forcing will be like is uncertain, since nobody knows how much CO2 will continue to be emitted by humans (or even by natural processes like volcanos). It is possible, however, to make what you might call educated guesses. Past emission trends can be easily continued into the future, and hypothetically modified depending on how quickly or slowly the world reduces its dependence on fossil fuels. One result of international collaboration is that a small number of such future forcing scenarios have been defined and are widely accepted as how climate models should be forced into the future. Each such scenario makes a different assumption as to what future greenhouse gas concentrations will be like.
So, starting from known past conditions, a climate model can run through the annual cycle for 30, 40, 100, or more years into the future using one of those standardized forcing scenarios. Each simulation into the future may use a different forcing scenario. Each simulation may also be run multiple times with the same forcing but with just slightly different initial conditions. An ensemble of sample climates can be generated like this. Each sample is equally valid (since they are all subject to the same forcing and boundary conditions), but thanks to the many nonlinear “chaotic” processes involved, each sample can also turn out to be quite different. Those differences provide a measure of how naturally variable the climate system can be. Furthermore, a large ensemble of future climate runs allows the uncertainty of the projections to be partially quantified. For example, if 99 out of 100 simulations respond in the same way to increased greenhouse gases, you can be “virtually certain” (in IPCC parlance) that the real world will respond similarly.
Climate modelling is a world-wide project
Over 30 different global climate models have been developed by different research centres around the world including Ireland via ICHEC and Met Éireann. Like the cars in a Formula-1 race, some of them share certain component parts, but each is also different in its own way. Their real strength, however, is when they are collectively organized to each run the same simulation experiments. Insofar as the particular design of each model permits, they all use the same initial configuration, the same boundary conditions, and the same forcing scenarios. Collectively, they generate ensembles of many thousands of sample climates for each forcing scenario. That number is big enough to generate high confidence that something close to the full range of possible climates is being captured. Multi-model ensembles like this are great examples of the whole being greater than the sum of its parts.
These coordinated climate simulation projects also reflect how climate research has become “big science”, with no one person or even one institute able to do it all. ICHEC’s simulations have consumed 10s of millions of computer core-hours and the resulting datasets occupy over 1.5 petabytes (equivalent to the storage capacity of 1500 well-endowed modern laptops).
Until the simulations are run, nobody knows for sure whether (or by how much) the primary forcing from higher concentrations of greenhouse gases will be amplified by positive feedbacks or damped by negative ones. Some feedback processes are very well understood and are clearly detectable, like the positive feedback between warming and ice cover. This works by some initial warming reducing ice cover, which allows more absorption of sunlight due to a decreased albedo, which in turn leads to yet more warming. Other feedbacks (especially those related to clouds) could conceivably go either way and so high-resolution simulations are needed to determine their most likely behaviour.
The potential for more extreme events
By now, virtually everybody knows that almost all models from over 30 years ago until now predict that the world will continue to get warmer if greenhouse gas emissions continue on their current trajectory. A mean warming of 2 or 3 degrees does not seem like much, given that temperatures can vary by a lot more than that just from day to day. However, even that amount of warming is likely to lead to widespread and even dramatic changes in ice cover (especially in the Arctic), to sea levels, and in the natural world of plants and animals.
Even more dramatic than changes in the mean climate are the projected future changes in the extremes, at the “tails” of the frequency distributions of each sample climate. Thus, a storm so violent that it only occurred once every 100 years or so in the past, may occur every 20 years in the future – or 5 times more frequently. Similarly, long deep freezes (such as occurred in Ireland during the winter of 2010-11) may become even rarer in a future warmer climate.
What does this mean for Ireland?
Researchers at ICHEC have supplemented the global climate simulations with a further series of regional simulations (nested within a larger global simulation) to provide a set of ultra-high resolution future climate scenarios for Ireland. The latest report (available from the EPA web-site at https://www.epa.ie/publications/research/climate-change/Research_Report_339_Part1.pdf ) contains a wealth of detail as to what future global warming scenarios are likely to mean for Ireland, from temperature, precipitation and wind changes to climate indices like length of the growing season and number of frost-free days.
Selected findings from this study indicate that by the middle of this century (2041–2060):
● Temperatures are projected to increase by 1–1.6°C compared with the baseline period (1981–2000), with the largest increases in the east;
● Warming will be enhanced at the extremes (i.e. hot days and cold nights), with summer daytime and winter night-time temperatures projected to increase by 1–2.4°C;
● Substantial decreases of approximately 50% are projected in the number of frost and ice days;
● Summer heatwave events are expected to occur more frequently, with the largest increases in the south;
● Precipitation is expected to become more variable, with substantial projected increases in the occurrence of both dry periods and heavy precipitation events
○ Projected increases in dry periods are largest for summer
○ Largest projected increases in heavy precipitation are expected for the winter and autumn months.
● Snowfall is projected to decrease substantially by the middle of the century with reductions of over 50%.
● The energy content of the 120 m (turbine height) wind is projected to decrease for all seasons with the largest decreases noted for summer.
● the length of the growing season is projected to increase by between 12% and 16%. Similarly, the grazing season, crop heat units and growing degree days for a range of crops (and pests) are projected to increase substantially by the middle of the century.
● Substantial changes in storm tracks, increases in mean sea level pressure, surface evapotranspiration, specific humidity and cooling demand, along with decreases in 10m wind speed, heating demand and solar energy resources are also projected by mid-century.
Climate researchers at ICHEC are working on improving the climate projection dataset for Ireland using the most up-to-date coupled atmosphere-ocean-wave regional climate models and updated emission scenarios for the period 1980-2100.
Supercomputing capability has increased by over 7 orders of magnitude, i.e., by a factor of 10 million, over the past 30 years. This has enabled advances in climate science which would have been unthinkable even as recently as 1990. With the vast computing resources at our disposal, climate researchers can provide very precise indicators of our climate. The level of detail and consistency achieved gives confidence in these projections and allows an ever more persuasive evidence-based consensus to emerge. The challenge now is for policymakers and individuals to respond with appropriate climate action.
- IPCC Glossary https://www.ipcc.ch/report/sr15/glossary/
- EC Earth Global Climate Simulations Ireland’s contribution to CMIP6 https://www.epa.ie/publications/research/climate-change/research-310-ec-earth-global-climate-simulations-irelands-contributions-to-cmip6.php
- AR1-6 IPCC https://www.ipcc.ch/assessment-report/ar5/