{"id":4477,"date":"2023-09-20T15:09:24","date_gmt":"2023-09-20T13:09:24","guid":{"rendered":"https:\/\/nersc.no\/?page_id=4477"},"modified":"2026-03-05T15:22:15","modified_gmt":"2026-03-05T14:22:15","slug":"artificial-intelligence","status":"publish","type":"page","link":"https:\/\/nersc.no\/en\/research-area\/models\/artificial-intelligence\/","title":{"rendered":"Artificial intelligence"},"content":{"rendered":"

[vc_section css=”.vc_custom_1669014075681{background-color: #ffffff !important;}” el_class=”full-image”][vc_row][vc_column]\n\t\t\t

\n\t\t\t\t\"\"\n\t\t\t\t\n\t\t\t<\/figure>[\/vc_column][\/vc_row][vc_row][vc_column width=”1\/6″][\/vc_column][vc_column width=”2\/3″]

Artificial intelligence<\/h1><\/div>

Computers can be programmed to analyze and recognize patterns in large amounts of data. This is called machine learning, an important field within artificial intelligence. In research, this is used, among other things, for the analysis of large amounts of data and for the forecasting of changes in the environment or climate. Through the use of artificial intelligence, we can prepare far better forecasts than before, and present good and secure analysis and decision-making bases.<\/p><\/div>[vc_column_text css=””]<\/p>\n

With the help of artificial intelligence, information can be extracted more efficiently and correctly from extremely large amounts of data, than with traditional statistical methods. Our work focuses on data from satellites, drifting buoys, and simulations from our models. Machine learning helps us to streamline analyses of data, and to discover new connections between the data. One example is the use of machine learning for faster classification of radar images of sea ice. We combine years of knowledge about the environment and climate with leading techniques for machine learning, and, through this, increase the quality of analyses and forecasts\/predictions. With the help of artificial intelligence, we will increase the speed of our models and data assimilation methods. Current studies are primarily aimed at operational sea-ice and climate prediction. The models will be able to learn from their mistakes, add missing details in a forecast or in satellite data, and for example predict specific types of harmful algal blooms.<\/p>\n

[\/vc_column_text]

<\/div>
\n

For more information, contact research leader Laurent Bertino<\/strong><\/a>.<\/p>\n<\/div>

<\/div>[\/vc_column][vc_column width=”1\/6″][\/vc_column][\/vc_row][\/vc_section][vc_row][vc_column]
<\/div>

News<\/h2><\/div>