Mikhail Krinitskiy gave a talk at the World Youth Festival about AI applications in climate science.

In his presentation at the panel discussion on climate change at the World Youth Festival, Senior Researcher Mikhail Krinitskiy discussed the application of Artificial Intelligence approaches in monitoring observations, measurements, and research tasks related to climate change. The talk highlighted the impact of climate change on weather patterns, species migration, and ecosystem dynamics, emphasizing the need for accurate monitoring and data analysis.

It is more accurate to talk about climate change rather than just global warming. While the average surface temperature of the ocean is indeed rising, and at a faster rate due to human influence, the effects observed by ordinary people are often short-term and may not necessarily manifest as warmer weather or drier seasons. In some regions, such as the Svalbard archipelago, it is even projected to become colder due to the current climate change.

The main effect of climate change in terms of atmospheric and ocean behavior can be characterized as making the weather "more nervous." The frequency and intensity of various extreme weather events are increasing. For example, the frequency and intensity of tropical cyclones are on the rise. There will be an increase in the number and intensity of mesoscale convective events over land, which are characterized by very strong winds and associated risks to infrastructure and even human life. Predictions suggest that precipitation on land will become more clustered in time and space, leading to an increase in the frequency and duration of floods and droughts. Climate change is already causing the migration of entire species of plants, animals, and insects, including pests or disease vectors, both plants and mammals, including humans, in some regions. This migration is not only about changes in habitat range for some species. During such processes, migrating species acquire the status of colonizers in new territories where, due to climate change, they may be better adapted to new conditions compared to native species of these ecosystems. Sometimes this can lead to the suppression or even complete displacement of native species.

Climate change also leads to changes in dietary habits, hunting, mating, and competitive behavior of species inhabiting territories where such changes are most pronounced: on the ice cover in the ocean, on the Arctic coast, in regions covered by permafrost.

Amidst all these observed phenomenological changes, the primary task facing researchers is monitoring indicators characterizing both the climate changes themselves and the causes that are known to lead to these changes. Projects such as Carbon Polygons (a pilot project of the Ministry of Science and Higher Education) and the Russian Climate Monitoring System (a key innovation project of national importance "Unified National System for Monitoring Climate-Active Substances") have been launched and are continuing to develop. Monitoring observations and measurements of meteorological parameters, climate-active gas fluxes are carried out on the sites of these projects, on land and at sea; methods for increasing the sequestration potential of territories or water areas are being developed.

In many studies conducted at the carbon polygon sites or within the Russian Climate Monitoring System projects, we are currently applying artificial intelligence approaches. For example, we have trained a neural network to determine carbon dioxide and methane fluxes from satellite imagery with higher accuracy compared to existing methods based on vegetation indices. We are also working on accelerating or improving the accuracy of ocean and atmosphere modeling in classical hydrodynamic models using neural network technologies. Additionally, we have applied an approach based on artificial neural networks to identify regions in Russia that are homogeneous in carbon cycle characteristics, allowing for more adequate planning of the placement of monitoring stations or carbon polygons.

It cannot be said that artificial intelligence is curbing climate change. However, in our research, where we conduct monitoring observations and measurements of parameters characterizing these changes, we actively employ artificial intelligence technologies today. This enables us to expand territorial coverage, increase measurement and modeling accuracy, or speed up research processes.

« Back