Credit: CC0 Public Domain

Over the past week we’ve witnessed the many political discussions that go with the territory of a COP—or, more verbosely, the Conference of the Parties to the United Nations Framework Convention on Climate Change.

COP30 is the latest event in annual meetings aiming to reach global agreement on how to address climate change. But political events such as COP base the need for action on available science—to understand recent changes and to predict the magnitude and impact of future change.

This information is provided through other international activities—such as regular assessment reports that are written by the Intergovernmental Panel on Climate Change (IPCC). These reports are based on the best available scientific knowledge.

But how exactly do they evaluate what will happen in the future?

Climate futures

Predictions of future climate change are based on several key planks of evidence. These include the fundamental physics of radiation in our atmosphere, the trends in observed climate and longer-term records of ancient climates.

But there is only one way to incorporate the complex feedbacks and dynamics required to make quantitative predictions. And that is by using climate models. Climate models use supercomputers to solve the complex equations needed to make climate projections.

The most sophisticated climate models are known as Earth system models. They ingest our knowledge of climate physics, radiation, chemistry, biology and fluid dynamics to simulate the evolution of the entire Earth system.

Climate centers from many different nations develop Earth system models, and contribute to a global data project known as CMIP—the Coupled Model Intercomparison Project. This data is then used by scientists worldwide to better understand the possible trajectories of, and to study the reasons for, future change.

Regional climate changes

Data from Earth system models cover the whole globe, but there is a catch. The computational expense of these models means that we run them at low resolution—that is, aggregating information onto grid boxes that are about 100 kilometers across. This puts the entirety of Melbourne, for example, within a single grid box.

But the climate information that we need to guide future adaptation needs more detailed information. For this, scientists use tools known as “downscaling,” or regional climate projections. These take the global projections and produce higher resolution information over a limited region.

This high-resolution information feeds into products such as the recently released National Climate Risk Assessment from the Australian Climate Service. Similar climate information is used by local governments, businesses and industry to understand their exposure to climate risk.

We’re doing it all again

Each iteration of CMIP, which began in 1995, has brought about improvements which have helped us to better understand our global climate.

For example, CMIP5 (from the late 2000s) helped us to understand carbon feedbacks and the predictability of the climate system. The CMIP6 generation of climate models (from the late 2010s) provided more accurate simulation of clouds and aerosols, and a wider set of possible future scenarios.

Now we are doing it all again—to create what will be known as CMIP7. Why would we do this?

The first reason is that more climate information has become available since CMIP6. CMIP simulations use “scenarios” to look at the range of plausible futures of climate change under different socio-economic and policy pathways.

For CMIP6, the “future” scenarios were started from the year 2015, using the information available at the time. We now have an extra decade of information to refine our projections.

The second reason is that CMIP7 shifts more to emissions-driven simulations for carbon dioxide, allowing models to calculate atmospheric concentrations on the fly.

Simulating how atmospheric carbon dioxide and other greenhouse gases interact with the land and ocean (known as the carbon cycle) allows feedbacks and potential tipping points to be calculated. However, this also requires a more complex Earth system model.

Australia’s CMIP7 contribution aims to incorporate new science and knowledge with a refined carbon cycle which includes Australian vegetation, bushfires, land use change and improved ocean biology.

Thirdly, this time around we aim to run models at higher resolution—such as having 16 grid boxes over Melbourne, instead of one. This is possible thanks to advances in computational capability and modeling software.

We’ve started the process

This week, Australia’s newest Earth system model version—known as ACCESS-ESM1.6—is initiating the first phase in the CMIP7 contribution process, which is supported through the National Environmental Science Program Climate Systems Hub.

This includes a long “preindustrial spinup,” where we run the model for about 1,000 virtual years using greenhouse gas levels from before the industrial revolution until the stable conditions are reached and available observations are matched. The spinup is required to ensure that all subsequent simulations start from a physically consistent state.

In the next phase, we’ll run a “historical” simulation that emulates the last 200 years of civilization. Only then can we implement a range of future scenarios and complete our climate projections.

This work is a partnership between CSIRO and Australia’s climate simulator (ACCESS-NRI), with support from university-based scientists and the Bureau of Meteorology. It’s an exercise that will take multiple years, consume hundreds of millions of compute hours on high performance supercomputers of the National Computational Infrastructure, and will produce about 8 petabytes of data—or 8 million gigabytes—to be processed and submitted to CMIP7.

As the only Southern Hemisphere nation submitting to past CMIPs, Australia has a unique and crucial perspective.

This data will also be used for higher resolution regional climate projections, which will then be used for future climate risk assessments and adaptation plans. It will also inform the IPCC’s next assessment report.

Ultimately, a future COP will translate this evidence into global action to further refine our climate targets.

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Citation: Behind every COP is a global data project that predicts Earth’s future—here’s how it works (2025, November 20) retrieved 20 November 2025 from https://phys.org/news/2025-11-cop-global-earth-future.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.