According to IAAR researcher Serguei Soldatenko, by applying artificial intelligence methods to observations of the past and present of the climate system, the intention is to create an application that can be used to predict the climate on a well advanced time scale.
In their work, specialists from IAAR and the Institute of Nature Management of the Belarusian National Academy of Sciences will focus on modeling the climate in the two countries over a 20-year time period.
Soldatenko explained that weather forecasting from months to 20 years in advance is a gray area in which there is a high proportion of uncertainty due to many factors that are difficult to predict.
Classical methods do not allow forecasts to be properly taken into account, however, self-learning digital systems, as scientists believe, should solve this problem.
“Our task is to create a completely new tool that does not deny everything that climatologists and mathematicians have already done, but allows us to obtain information about future climate that current methods do not provide,” Soldatenko emphasized.