Predicting climate change

With many different factors affecting climate change, how can we predict the climate with any accuracy?

The Natural Environment Research Council co-funds the world's largest climate change prediction experiment Climateprediction.net. The experiment runs a computer-based model of our climate thousands of times, each with slightly modified starting conditions. This lets researchers thoroughly investigate the range of likely changes. Details are available on the Climate Prediction website.

UK scientists are developing and testing other state-of-the-art climate models. For example, the HiGEM project will develop one of the world's highest resolution global Earth system models. The UK and Japan are collaborating to run this high-resolution model on one of the world's biggest computers. Details are available on the Earth Simulator website.

Climate Models

What is a climate model?
How do models help us understand climate?
Are climate models accurate?
Are climate models too simple?

What is a climate model?

Climate models are based on mathematical equations that describe the behaviour of the atmosphere and ocean, and their interactions with other components of the Earth system, such as the cryosphere (ice sheets and glaciers), the land surface and the biosphere (living plants animals and microbes).

A model's core equations, which are derived from the laws of physics, describe how temperature, pressure and winds (or currents) and other variables in the atmosphere and ocean change over time. Additional equations describe chemical and biological aspects of the climate system.

In the model, the climate variables are represented on a three-dimensional grid covering the atmosphere and the oceans. The spacing between grid points in the atmosphere is typically a few hundred kilometres horizontally and 500m vertically.

How do climate models help us understand climate?

The climate system is very complex, and understanding it is difficult. To make progress, we need to explore mechanisms and test theories. It is not feasible to experiment on the climate system itself, nor is it possible to reproduce the full complexity of the climate system in a laboratory. Climate models offer us the best possible alternative - a numerical laboratory where we can address vital questions.

How will the climate change in response to rising levels of greenhouse gases?
What would happen to the climate of Europe if the Atlantic Ocean's thermohaline circulation (the current that carries warm water to northwestern Europe) stopped?
What controlled Earth's climate in the past?

Are climate models accurate?

Climate researchers are confident that climate models can accurately represent key aspects of the climate system.

Climate models are continually evaluated against datasets of real observations. This shows that the models can reproduce many aspects of present and past climate, including the overall strength and pattern of recent climate change researchers have monitored.
Climate models can successfully forecast the weather, and even major climate phenomena such as El Niño, an important climate feature in the Pacific Ocean.
The core of a climate model uses well-understood physical, chemical and biological equations and principles.

Climate models are the only scientifically-credible tools for making detailed predictions about climate at the scale of geographical regions. Nonetheless, because climate models are mathematical approximations of the climate system and not the real system itself, their results must be treated with due scientific caution.

Are climate models too simple?

Climate models include as many physical, chemical and biological processes as possible. Current models are certainly good enough to simulate large-scale climate phenomena, and researchers continually check that models do this well. Identifying the limitations of a particular model is an important part of stimulating further improvements, and advancing our understanding.

Found this page useful?

About social bookmarking