Climate change brought home

The planet is warming, but how will this affect local weather? Christian Jakob, professor of climate modelling at Monash University, is working on the answers.

Story Bianca Nogrady

What’s the key driver behind your research?

The international community, under the Paris Agreement, is trying to hold global warming to within 2ºC. But that’s a global mean temperature. People want to know what’s going to happen to their local weather – in Melbourne, in London, wherever they live.

What’s the biggest challenge in assessing climate change impacts on weather locally?

Predicting rainfall. Everybody’s interested in rainfall, but it’s hard to predict.

When you’re talking about climate in a location or small area, everything depends on the atmospheric and oceanic circulation, and the questions of how the circulation is going to change and how this will affect local weather is a hard one to answer with our models. Even weather forecasting is pretty bad at predicting rainfall.

What’s your research group doing to improve rainfall predictions?

We compare actual observations with what we can learn from process models, which explicitly represent the individual processes involved in weather events, such as the physics of thunderstorms. These process models are more sophisticated than current weather and climate models, but due to computational constraints we can run them over small areas only. By analysing both observations and process models, we identify where the biggest contributions to error in our prediction models are coming from, which may be, for example, how we represent thunderstorms in the model. Once we know the source of the errors,we can make targeted improvements. For example, in one project we’re improving regional predictions of rainfall over Australia by addressing a modelling error that overestimates rainfall over the tropical Indian Ocean.

What have been the key advances in climate science and climate modelling in recent years?

Over the past three decades we’ve improved our ability to predict weather by about one day per decade – a three-day forecast today is as reliable as a one-day forecast 20 years ago. In climate modelling our focus has been on increasing the complexity of the models so they can do more. For instance, we’ve been able to add the interaction of vegetation with the atmosphere. We’ve also added additional processes such as atmospheric chemistry and biogeochemistry of the ocean, so we can model the carbon exchange with the ocean. We’ve also increased the resolution of climate models, which increases their level of detail and their accuracy.

What impact do you hope your research will have on communities’ ability to adapt to a changing climate?

Rain is essential for agriculture, it’s essential for food security, and it’s essential for water security, so predicting changes in rainfall – whether that’s next winter or next summer or over the next 50 years – is really important. And we’re not just interested in ‘average’ rainfall changes, but how it falls. Will we have more big rain events causing floods, or will it be a more gradual increase?

I’m working with a group of farmers in Australia called the Managing Climate Variability Program on research to improve regional rainfall predictions. People are interested in rainfall, even if it’s just because a barbecue is being planned.

Almost every decision we make that’s weather or climate-related is now based on a prediction, and every prediction we make is based on models. So if we want to make better decisions, we need to make better predictions, and to make better predictions we need better models.