The "Yandex Weather" service has implemented OmniCast technology. It receives data from amateur and professional weather stations 36 times more frequently and creates a more accurate weather picture for each city block, the company's press service reported.
The "Yandex Weather" service is based on Meteum technology, which uses information from various weather stations (satellite and ground-based). However, they are all located unevenly. Therefore, it was not possible to give an accurate forecast where there were few of them. The OmniCast neural network solved this problem.
OmniCast fills this gap with amateur weather stations. They are located everywhere and transmit information every five minutes, while professional ones do so every three hours. This allowed us to receive data more often and recalculate the forecast.
OmniCast is powered by artificial intelligence. The technology takes into account the Meteum forecast, air temperature fluctuations, and weather station data. This makes it possible to make accurate forecasts for each area of 500 by 500 meters. Local forecasts are especially relevant for large cities.
Earlier, www1.ru reported that Yandex created a neural network to detect rare pathologies in early pregnancy.
Read materials on the topic:
Spell check, data analysis: "Yandex" launched the YandexGPT 3 Lite neural network