ST. PETERSBURG, October 22. /TASS/. The highest water levels in the northern rivers during the spring flood will be forecasted using artificial intelligence. The long-term forecasting method was developed by scientists of the Arctic and Antarctic Research Institute (AARI, St. Petersburg), the institution's press service said.
"Traditional methods of forecasting spring floods rely on snow reserves data. However, in the Arctic regions, in early March, this information remains insufficient for early forecasts: the snow accumulation still continues there. The sparse observation network and snow transport by snowstorms add to the difficulties, and due to this, individual measurements' results cannot always reflect correctly the situation in the entire area. When developing new methods of forecasting spring water rise levels, the scientists additionally took into account effects of autumn moisture and winter soil freezing," the press service said.
What additionally affects the water levels in spring is the autumn weather, including how intensive rains are and at what rate marshes are freezing up.
"In the tundra, rainy autumn and swamps, overflowing with moisture, turn the earth's surface layer into a solid ice shield during winter freezing. Further on, in spring, during the snowmelt period, the frozen soil absorbs meltwater much worse, and thus the water enters river channels in bigger amounts. It is this feature that contributes to the formation of high floods in Russia's northern regions. In case of dry autumns, the soil can absorb more meltwater, which cuts the risk of a sharp rise in river levels," the press service explained.
According to institute's representatives, in the control area - on rivers of the Yamalo-Nenets Autonomous Region - the accuracy of forecasts that use the new methods was more than 80%, which is almost twice higher than regular climate forecasts have.
Practical application
The research, aimed at developing long-term hydrometeorology forecast methods for rivers and seas of the Russian Arctic in the changing climate, was conducted by AARI specialists for further use at the Ob-Irtysh Department of the Hydrometeorology and Environmental Monitoring.
The scientists have created a system of 120 physical features to build digital models for rivers of the Yamalo-Nenets Autonomous Region. The system includes rates of groundwater supply depletion, changes in water levels, accumulation and redistribution of water reserves, as well as dynamics of hydrological processes.
The algorithms were adjusted taking into account specifics of every basin, including for the Nadym, the Pur, the Taz, and the Pyaku-Pur Rivers.
"Basins of the Nadym, the Pur and the Taz Rivers are of strategic importance for the Yamalo-Nenets Autonomous Region's oil and gas industry. Here are big deposits, compressor stations, and main gas pipelines, therefore extreme floods may pose a serious threat to the region's industrial and transport infrastructures. An important feature of northern rivers is that they flow mainly from south to north. Thus, in spring, meltwater rushes to the estuaries, where the ice cover may still remain thick and durable. This creates conditions for the formation of ice jams that obstruct the water flow and lead to a sharp increase in its level," senior researcher at AARI's Department of Hydrology of Estuaries and Water Resources Nadezhda Volkova said.
According to her, the developed methods will be used to assess well in advance risks of dangerous flood developments to plan preventive measures.