Computer simulations that scientists use to understand the evolution of the Earth’s climate offer a wealth of information to public officials and corporations planning for the future. However, climate models contain some degree of uncertainty, no matter how complex or computationally intensive. As decision-makers are asking more complex questions and looking to smaller scales, addressing this uncertainty is proving increasingly important.
To improve climate simulations, scientists are looking to the potential of Artificial Intelligence (AI). AI has offered profound insights in fields from materials science to manufacturing, and climate researchers are excited to explore how AI can be used to revolutionise how the Earth system, and especially its water cycle, can be simulated to dramatically improve our understanding and representation of the real world.
In particular, AI offers the potential to dramatically increase the accuracy of predictions down to the scales of interest to scientists, and even stakeholders focused on designing, financing and deploying equitable climate solutions to America’s most disadvantaged communities.
Earth system predictability refers to the intersection of climate with hydrology, ecology, infrastructure and human activities. AI for the climate is still in its infancy, however, it is still essential that we explore the potential of AI to see how it can better inform our models and prepare us for the future.
– Nicki Hickmon, Argonne Scientist, Director of Operations, Atmospheric Radiation Measurement User Facility
Motivated by this opportunity, the U.S. Department of Energy (DOE) is launching a comprehensive workshop: Artificial Intelligence for Earth System Predictability (AI4ESP). After the collection of more than 150 white papers from the scientific community, AI4ESP is kicking into high gear by hosting a workshop. The workshop will include 17 sessions over a six-week period designed to create a new scientific community that marries climate research with AI, applied math and supercomputing.
By linking researchers in Earth system predictability and computer sciences, AI4ESP seeks to create a paradigm shift in simulating the Earth system. AI4ESP seeks to inspire a new generation of AI algorithms specifically aimed at Earth system predictability. Continuous improvements will enhance the ability of current simulations to provide deeper insights into community-scale issues and those involving extreme weather, potentially allowing stakeholders a better grasp of the uncertainties that surround such events.
A paper titled “Tackling Climate Change with Machine Learning” offers up 13 areas where machine learning can be deployed, including energy production, CO2 removal, education, solar geoengineering, and finance. Within these fields, the possibilities include more energy-efficient buildings, creating new low-carbon materials, better monitoring of deforestation, and greener transportation.
As reported by OpenGov Asia, experts are now using machine learning to help solve one of humanity’s biggest problems: climate change. With machine learning, researchers can use the abundance of historical climate data and observations to improve predictions of Earth’s future climate. These predictions will have a major role in lessening our climate impact in the years ahead.
Machine learning algorithms use available data sets to develop a model. This model can then make predictions based on new data that were not part of the original data set. Regarding climate change, there are two main approaches by which machine learning can help further the understanding of climate: observations and modelling. In recent years, the amount of available data from observation and climate models has grown exponentially. Hence, machines can analyse all of the data.
As the computational capacity grows — along with the climate data — researchers will be able to engage increasingly sophisticated machine learning algorithms to sift through this information and deliver improved climate models and projections.
