Role of Artificial Intelligence in the Battle Against Climate Change

As an old Native American saying goes, “When the last tree has been cut down, the last fish caught, the last river poisoned, only then will we realize that one cannot eat money.” We are at the brink of a climate crisis. According to climate scientists, the earth is heating up at an unprecedented rate, thanks to all the human industrialization in the last hundred years. While industrialization has brought on the many benefits and comforts that we know of today, its consequences will soon overpower the pros if we do not act fast enough. According to the Paris Agreement, humanity must limit the increase in global average temperature to 1.5° C above pre-industrial levels to prevent the harsh climate impacts in the years to come.

Modern technological advancements, which has also been a boon of the industrial revolution has brought forth a multitude of advantages, which has been shaping the world we live in today. One such advancement of technology is Artificial intelligence. AI, a term first coined in 1956, has been around for a long time but recently with the development of large language learning models (LLM) like ChatGPT, ClaudeAI, Deepseek etc have once again come to the forefront of public attention. As we currently witness the various domains in which AI is being implemented including business, healthcare, education and many more, we need to keep in mind that it can also be a tool which has immense potential in equipping us in the fight against climate change.

In recent times, the UNFCCC has taken an initiative named Artificial Intelligence for Climate Action Initiative which seeks to ensure that the international community leverages this tech to uphold the guidelines set by the Paris agreement.

Artificial intelligence has the ability to  accelerate climate research, forecast climate disasters, monitor remote sensing and increase climate literacy among numerous other solutions.

Before that, let’s start with the basics. What do we mean by artificial intelligence? Can a machine really be intelligent on its own?

For starters, AI systems are created by feeding massive amounts of data into a computer program. The program is then trained to learn patterns and correlations in the data, and when new data is passed to the system to solve a problem, it mimics those patterns to generate a solution, in the form of a prediction, classification or other. This allows AI to reason, learn, and self-improve. This ‘learning’ or ‘intelligence’ is what sets AI apart from other technologies.

In the context of climate change, this ability holds immense power to revolutionize industries. Despite the general consensus about the basics of climate change, there is still huge debate and uncertainty on certain scientific aspects. This is where AI comes into play. AI, being trained on millions of multi-dimensional sets of data, is the best bet in making the most likely predictions of future trends and outcomes of different climate scenarios.

AI can learn from the past data, test its accuracy, and make predictions about the future.

During the last decade, AI has been used to predict and forecast many climate events. A paper published in Remote Sensing (2021) states that “AI will become an indispensable part of state-of-the-art weather forecasting and climate monitoring and prediction” In another study published in Frontiers in Robotics and AI in 2019, the global mean temperature was predicted with an accuracy of 97% using neural networks.

AI in Agriculture and Food

Climate-smart technology is revolutionizing the field of agriculture by aiding farmers in making informed decisions by providing them with real-time weather forecasts, data analytics of yields, and resources which they otherwise would not have access to. For instance, in rural areas, farmers can upload images of their crops on their phones, and using image processing and AI technology, plant diseases can be detected without the need for consultation with experts. These practices help farmers to be more climate resilient and get the best harvests, boosting food production. For example,  IGAD Climate Prediction and Applications Centre is a useful resource for farmers, climate activists, and scientists in Africa which uses AI in agricultural forecasts and forecasting cyclones, droughts, and weather patterns in East Africa.

AI in Power Grid

As the world slowly transfers to clean energy, a lot can be done in the meantime to reduce the climate impact of traditional fossil fuel industries. Power grid i.e transformers that supply electricity from power plants to our homes can be optimized using ML models that are trained to understand the supply and demand of electricity and switch between different renewable and non-renewable sources depending on the demand.

AI in Optimizing Supply Chains

The supply chain process can also  be automated and optimized by this technology. By investing in AI based automation and optimization processes, companies can make greater margins of profit while also reducing the emission of greenhouse gasses. ML can help reduce emissions in supply chains by intelligently predicting supply and demand, identifying lower-carbon products, and optimizing shipping routes, saving fossil fuels.

AI for Environmental Monitoring and Conservation

In today’s fast-paced world, AI can play a crucial role in providing policy makers with data driven insights and aid them in making climate decisions. AI-powered sensors and drones are also used to monitor air and water quality, helping researchers and policymakers respond quickly to environmental hazards. For example, semantic segmentation of satellite images of the Amazon rainforest is used to monitor the forest cover change in the Amazon.

Challenges and Risks

On the other hand, to keep up with the ever growing computational demands in this field, AI companies are opening data centers all over the world. Data centers are facilities which are equipped with huge computing resources such as computers with high performance GPUs which provide the equipment needed to do the computations for training and running deep learning models. These machines use huge amounts of electricity, hence contributing to global carbon dioxide emissions. According to the paper “Energy efficiency in cloud computing data centers: a survey on software technologies” published in Cluster Computing (2023), the energy consumption of data centres is predicted to rise from 200 TWh in 2016 to 2967 TWh in 2030. AI itself has a carbon footprint of its own and is not zero emission. Although, to tackle this issue, Google’s Deep Mind is creating algorithms that optimize computer systems, make computations more efficient and design more efficient chips.

But the issue of ethical consideration and transparency still remains. While solutions provided by neural networks and black box algorithms may seem perfect at first, it is difficult to explain how or why they arrived at that particular solution because of their complexity. Therefore it is crucial that important decisions be made in conjunction with human experts.

Like most technologies, these reasons make AI a double edged sword that we must use with caution. Interdisciplinary collaboration and human regulation is essential for ethical development of this technology and the welfare of society.

About the Author:

Tausia Tahsin Nuzum works as a research Intern at ICCCAD.

The views and opinions expressed in this blog are solely those of the author. This write-up is a personal reflection and does not represent the official stance of ICCCAD.