Skip to content

Response: predicting the future – climate models

Predicting the future: climate models

(Image: Caltech)

Predicting the future : climate models

  • Earth systems that affect and are in turn affected by climate are complex and dynamic. A small change in one such as the temperature of an ocean current leads to a cascade of changes and feedback effects.
  • Climate models are driven by questions such as what climate was like in the past, what causes climate change, and what the climate might be like in the future, globally, regionally, and locally. This helps us understand the effects and the actions needed to avoid the worst impacts.
  • Models are limited by computing power, the volume and quality of available data (often difficult to obtain in places like Antarctica), the scale of what’s being modelled, and the complexity and unpredictability of chaotic non-linear events such as tipping points. Coupled models (Fig. 1) produce more realistic outcomes.
  • The bulk of the data used to model current predictions is from the Northern Hemisphere, yet what happens in Antarctica and the Southern Ocean has a profound influence in the rest of the world. Modelling these complexities requires some of the world’s largest supercomputers to run multiple iterations, but Zealand doesn’t have the resources, so it’s working in partnerships with the Unified Model Consortium, led by the UK Met Office to produce The New Zealand Earth System Model (NZESM)(Fig. 2).
Fig. 1: (Image: Carbon Brief)
Fig. 2: To “run” a model, scientists divide the planet into a 3-dimensional grid, apply the basic equations, and evaluate the results. Atmospheric models calculate winds, heat transfer, radiation, relative humidity, and surface hydrology within each grid and evaluate interactions with neighbouring points. (Image: NOAA)
Fig. 3: “The New Zealand Earth System Model (NZESM) couples together representations of atmospheric physics (wind, temperature and water in the atmosphere and the processes that link them), ocean dynamics (oceanic temperatures, currents and salinity), sea ice (both sea ice coverage and sea ice thickness), and land physics (soil moisture, soil temperature and river run-off). The model also represents some chemical, biological and land-ice aspects of the “earth system”: chemistry of the lower and middle atmosphere (with a focus on ozone), ocean ‘biogeochemistry’ (think plankton and dissolved carbon), and Antarctic ice shelves.”  – Deep South Challenge, New Zealand.

Models used by the IPCC

Since 1992, several models to predict future scenarios have been used by the IPCC, each improving on previous versions.

Interpretting IPCC graphs

Emission scenarios used in earlier IPCC reports.

These considered what the climate would be like based on different scenarios: how much greenhouse gasses we might emit and what priorities we gave to the environment vs the economy, called ‘Representative Concentration Pathways’ or ‘RCPs’ (see Fig 9. for a more detailed explanation).

These pathways were presented on graphs as an alphanumeric code: A2,  AB1, etc (Fig. 4). Figure 9 explains what these mean, and why different emissions pathway ‘storylines’ were considered.

Fig. 4: Representative Concentration Pathways (RCPs) (Image: IPCC 2007)

In the Fifth Assessment Report 2013/2014, Representative Concentration Pathways (RCPs):

The represent the concentration of greenhouse gasses in the atmosphere based on how these gasses retain heat.

  • Heat is measured in watts per metre squared, written as W⋅m2
  • In most graphs, the numbers 2.6, 4.5, 6.0, and 8.5 are W⋅m2 however W⋅m2 is implied, and the four units are written as four scenarios: RCP2.6 etc. (for example, Fig. 5).
  • In some graphs, this is written without a decimal place: RCP26, RCP45 etc.
Fig. 5: Representative Concentration Pathways (RCPs) (Image: IPCC)

Some graphs show the Representative Concentration Pathways (RCPs) over time, based on the emissions of fossil fuels. The coloured lines in Figure 6, for example, show dozens of model runs for the different RCPs and the potential temperature range for each scenario. Temperatures are shown as a range rather than an exact figure, because complex feedback effects and the temperatures at which irreversible tipping points are breached is uncertain.

Note: these are average global temperatures; the tropics are not warming as rapidly while the Arctic is warming much faster.

Fig. 6: Representative Concentration Pathways (RCPs) (Image: Global Carbon Project)

Shortcomings and criticisms

The complexity of climate change, the careful pace at which research is reviewed by other scientists before being published, the assumptions that governments would act jointly to reduce carbon emissions, and the limitations of modelling complex global systems about which we still have only limited knowledge, means that models are limited in terms of how well they can predict the future.

This was certainly the case for the 2007 Fourth Assessment Report (AR4). The modelling for sea-level rise, for example, was based in part on NIWA generated reports published six years earlier in 2001,  themselves based in part on 2001 IPCC projections based on twentieth-century understanding (or lack) of how quickly ice sheets could melt.

The worst-case scenario, A1F1 (see Fig 8. for a detailed explanation), was considered least likely, in part because it was assumed that  governments would act to reduce carbon emissions.

At the 2009 Climate Change Conference (Copenhagen Summit or COP15), unequivocal evidence was presented showing that the worst-case scenario for rising sea levels were being met or exceeded (Fig. 7), while Arctic Sea ice loss (see here for why this is so important) was literally falling off the charts (Fig. 8).

The Summit concluded that unless governments acted to reduce carbon emissions the planet is committed to temperature increases of 3–7°C by 2100, with 7°C most likely if the projected temperatures match the upper limit of A1F1 sea-level predictions.

In defence of the models that resulted in these graphs, the AR4 IPCC report clearly stated that these sea level predictions largely excluded contributions from melting ice caps because ‘the physics of ice sheet dynamics was not sufficiently well understood’.

In short, they included a disclaimer, one that was almost universally ignored in the media and by policy makers.

The basis for higher projections of global mean sea level rise in the 21st century has been considered and it has been concluded that there is currently insufficient evidence to evaluate the probability of specific levels above the assessed likely range.”  IPCC 2013

Fig. 7: Global (eustatic) sea level change 1970-2010. The red line is from observations measurements taken from tide gauge data; the blue line from satellite data. The grey band shows the projections of the IPCC Third Assessment report, the top of which was the ‘worst case scenario’. In short, observed sea level rise was matching the ‘worst case scenario’ model. (Image: Copenhagen Diagnosis p37).
Fig. 8: The blue area represents the predicted amount of sea ice in the Arctic each September (when sea ice is at its minimum), based on thirteen IPCC AR4 models. The solid black line in the centre is the mean of the combined models. The red line is what was actually observed to have happened (Image: Copenhagen Diagnosis p30).

Current models

The current climate models used by the IPCC (2013/2014), which are in turn used by the New Zealand Ministry of the Environment, and the basis for the 2015 Paris Climate Accord are also based on assumptions that have not come to pass:

And crucially:

A new generation of climate models

“The New Zealand Earth System Model (NZESM) is a state-of-the-art modelling system that couples together representations of atmospheric physics (wind, temperature and water in the atmosphere and the processes that link them), ocean dynamics (oceanic temperatures, currents and salinity), sea ice (both sea ice coverage and sea ice thickness), and land physics (soil moisture, soil temperature and river run-off).

“In addition, the model represents some chemical, biological and land-ice aspects of the “earth system”: chemistry of the lower and middle atmosphere (with a focus on ozone), ocean “biogeochemistry” (think plankton and dissolved carbon), and Antarctic ice shelves.

“[The new] models will contribute to the upcoming 6th Coupled Model Intercomparison Project (CMIP6), which will inform the Sixth Assessment Report of the IPCC.” – National Science Challenges NZ

Fig. 9: Pages 4-5 in the 2000 IPCC ‘Summary for Policy Makers Emissions Scenarios’ explains why different emissions pathway ‘storylines’ were considered, and how they apply to the 2001 and 2007 IPCC reports. Note that the ‘A1F1’ is the ‘worst case scenario’.


The physics and chemistry:

Of Earth’s climate is well understood. Climate models are designed based on fundamental laws of physics, including the first law of thermodynamics and the Stefan-Boltzmann Law, which explains how natural greenhouse gasses (natural forcings) keeps the Earth’s surface ~33°C warmer than it would be without them.

Other mathematical equations are used to describe climate dynamics, including the Clausius-Clapeyron equation (the relationship between air temperature of the air and its maximum water vapour pressure) and the Navier-Stokes equations of fluid motion (speed, pressure, temperature and density of  gases in the atmosphere and the water in the ocean).

Video 1 explains in simple, clear terms how greenhouse gasses work and why more of them make the Earth warmer. For more about greenhouse gasses, click here (page on this website).

Video 1: How greenhouse gasses work

New Zealand Earth System Model (NZESM):

The UK Met Office Hadley Centre use the same Unified Model for both weather and climate modeling, using three Cray XC40 supercomputers capable of 14,000 trillion calculations per second. This system enables scientific and technical collaboration on a shared modelling system.

One new tool built to run and monitor complex modelling suites is Cylc, originally developed at NIWA, now an Open Source collaboration involving NIWA, Met Office, and others.

Carbon capture and storage technologies (CCS/BECC etc): see page this website. Some of these systems advocate for replacing native ecosystesm.

A much simpler, cheaper, and large scale solution is to restore biodiversity.

  • On the land, living plants store carbon in their leaves and trunks, while healthy native biodiversity create layers of rich soil that support micro-organisms that store carbon underground.
  • In the oceans, kelp stores huge amounts of carbon and grows faster than forests, as too does short-lived plankton than falls to the deep ocean floor. See the carbon cycle (this website).

Recognising the peril:

Even if countries act on their Paris climate agreement pledges to reduce emissions so that temperatures do not go beyond 1.5°C, the most recent UN Report states that we are on track for warming of more than 3°C by the end of the century (Video 2).

Video 2: The Emissions Gap

References and further reading