The role of AI in fighting climate change

March 30, 2020
Hot topics 🔥
AI & ML InsightsSustainability Trends
Mario Grunitz
The role of AI in fighting climate change

Find out everything you need to know about artificial intelligence and automation, and their applications in the working world with our new book, Working Machines – An Executive’s Guide to AI and Intelligent Automation. It is available on Amazon, Google Books, Apple Books and Audible.

All life on earth is in the midst of facing the most detrimental challenge to its survival — climate change. Rising sea levels, natural disasters multiplying more frequently, and the destruction of millenia-stretching ecosystems; the effects of climate change present severe consequences to every biological organism on the planet. While some do not share the same appreciation for the current and looming severity of the situation, others are making it their mission to explore innovative new ways to tackle the effects of climate change. The thing about science is that it exists unequivocally, whether you believe it or not.

Climate change scientists are frantically searching for effective approaches to fight against the disastrous effects of climate change.A novel yet unsurprising way they are doing this is tapping into the far-reaching capabilities of artificial intelligence, specifically machine learning.

So how do lines of digital code and complex algorithms possess the power to help save all of earth’s biological matter from total destruction?

Binary for biology

In November 2019, a group of the world’s most prominent machine learning experts, including minds from leading universities, Google, Microsoft, and DeepMind, released a detailed paper entitled Tackling Climate Change with Machine Learning. The group of volunteers, who work together under the title Climate Change AI, have detailed how machine learning’s capabilities can help by enabling or accelerating various strategies to fight against climate change. They argue how machine learning can be instrumental in reducing greenhouse gas emissions and in helping to identify major gaps that can be filled by this technology.

Below we have summarised a few machine learning solutions the Climate Change AI team identify in their report. If you want more detail (and boy do they have it) be sure to read the full report here.

Predictive capabilities

Improved predictions will be invaluable in helping officials develop more informed climate policies to allow governments to prepare for change, and mitigate against perpetuating future climate change. In order to rely on renewable sources of energy, we first need to be able to accurately predict the amounts of energy required on a regional and global scale. Through real-time analysis, policymakers are better informed to make both current and future decisions based on up-to-date data provided by machine learning algorithms. Essentially, the role of machine learning in this area will focus on helping us to build better electricity systems. By forecasting the immense amount of data gathered from global electrical systems on demand, scientists are now able to inform energy suppliers on how to incorporate renewable energy sources into regional grids to reduce waste.

Machine learning also helps reconstruct previous climate conditions using large-scale models to predict weather regionally, as well as monitoring the socio-economic impacts of both climate and weather. Machine learning’s predictive capabilities are now giving scientists more time to examine? Pending climatic issues and adapt to them as and when they occur.

Invaluable insights

Machine learning has the capability to draw invaluable insights from tons of complex data mined from climate modelling. Climate models help predict the change of average conditions of a region over time using quantitative methods to simulate the factors driving climate, including oceans, atmosphere, ice and land surface. Machine learning helps gather and synthesise this overwhelmingly large amount of data to help us gain more accurate indications of global carbon emissions on demand.

Previously, the issue with climate modelling was that although its simulations presented accurate data for short-term, its long-term predictive capabilities were fundamentally flawed. But that is a thing of the past thanks to machine learning, as we are now able to predict long-term climate change effects more accurately than ever before. When you have a basic idea of where you are headed you are limited in your decision making, but as soon as you get a much clearer picture of the future you can make informed and accurate decisions. This is made possible by machine learning.

Carbon emissions modelling

Machine learning can analyse and automate satellite images of the world’s power plants to get real-time updates on global emission levels. Now corporations and industries are no longer in control over the amount of information the public receives about their carbon emissions, which opens the door to wider accountability and ownership. This is a huge step towards more climate responsibility. From this, we are able to identify the major culprits of global carbon emissions and apply pressure to hold them accountable by imposing laws and fines if they exceed the accepted amount of emissions. It also gives us new ways to measure an individual plant’s carbon footprint and impact on the region by gathering and interpreting data of nearby infrastructure and electricity use.

Helping to create green materials

The Climate Change AI group states that close to 9% of the world’s greenhouse gas emissions derive from our buildings — more specifically the materials used. The production of steel and concrete has long created a large emission footprint for the infrastructure development industry, and it is time we begin sourcing alternative sources and low-carbon materials.

Machine learning is helping us develop low-carbon alternatives to these materials. How? Scientists are now able to model the properties and interactions of new chemical compounds in an ongoing experiment effort to find low-carbon materials. The world’s population is climbing and we’re going to need to develop more sustainable and green infrastructure to accommodate for both the rise in population and the effects of climate change in the future. Thanks to machine learning the path to discovery has been accelerated tremendously.

Efficient transport

Global transportation is responsible for nearly 25% of all the world’s carbon emissions, with a majority of this figure created by general road users. Machine learning can bring about more precise efficiency by helping to identify and minimise the number of wasted journeys, and improving vehicle efficiency through clever design. Another way machine learning is helping fight the CO2 emissions within the transport sector is by helping us switch freight carriers to more carbon-friendly options like railways. Thanks to Elon Musk and machine learning, we will soon be reducing the usage of personal cars through the development of shared, self-driving cars.

The import/export industries which rely heavily on the shipment of goods around the world is another major culprit of CO2 emissions. These processes involve all types of carbon-heavy transportation, from ships and aeroplanes to trucks and cars. Machine learning is finding ways to streamline transportation by ensuring multiple shipments are bundled together to reduce the number of trips. Think of it as the Uber XL of global logistics — fewer vehicles on the roads, in the air and on the sea means fewer emissions — and we all arrive alive together.

Green infrastructure

If you are wondering where the next large portion of the world’s greenhouse gas emissions comes from — it’s our buildings. Yep, the very place you might be reading this from. The amount of energy used in buildings around the world amounts to roughly 25% of our global CO2 emissions — that’s lights, heaters, air conditioners, kettles, computers, etc. — basically anything that relies on electricity that is commonly found in your average building around the world.

The wonderfully innovative capabilities of machine learning is helping us to create smart-buildings where sensors can monitor energy use such as water and air temperature, all from a device or application. A few minor adjustments can be made on existing buildings to help reduce energy waste. AI technology can help identify when applications are not being used and put them into hibernation. If this is done the way scientists predict, we will be able to reduce the average building’s emissions by roughly 20% — just think about the possibilities if we are able to roll this out on a city-wide scale.


Machine learning has a clear and profound role in humanity’s greatest fight for its survival. As it currently exists, there are a plethora of exciting and innovative ways machine learning can be incorporated into the fight against climate change — and as the technology improves so will its appetite for discovering new ways it can be useful. The most important thing to remember is that machine learning will work best when used in collaboration with various fields and infrastructures, including global business and political galvanisation. As the Climate Change AI group reminds us, machine learning’s role in the fight against climate change is no silver bullet as it relies on the backing of political action. But as it currently stands it is our only surefire way to start the fight with our best foot forward.

Mario Grunitz

Mario is a Strategy Lead and Co-founder of WeAreBrain, bringing over 20 years of rich and diverse experience in the technology sector. His passion for creating meaningful change through technology has positioned him as a thought leader and trusted advisor in the tech community, pushing the boundaries of digital innovation and shaping the future of AI.

Working Machines

An executive’s guide to AI and Intelligent Automation. Working Machines takes a look at how the renewed vigour for the development of Artificial Intelligence and Intelligent Automation technology has begun to change how businesses operate.