MOSCOW (PRWEB) May 22, 2019
Yandex, a technology company that builds intelligent products and services powered by machine learning, has greatly extended the reach of hyperlocal nowcasting in Russia's largest weather forecaster, Yandex.Weather. After adding data from a second satellite, nowcasting is now available for nearly all 53 million monthly Yandex.Weather users in Russia, expanding from its initial rollout in European Russia.
Yandex.Weather incorporates deep learning AI technologies to provide users with reliable weather reports. Yandex's in-house Meteum forecasting tech, the backbone of Yandex.Weather, combines traditional weather models with machine learning to produce highly accurate weather predictions. Yandex.Weather became even more precise in 2017 after the introduction of hyperlocal nowcasting, providing minute-by-minute precipitation predictions within a two-hour window based on a user’s location.
Yandex has also continued to improve how people can access Yandex.Weather and ensure it is always available in the most convenient ways possible. Over the past few years, as the Yandex ecosystem of products and services has grown, Yandex.Weather has grown with it. One of the most popular uses of Yandex.Weather is through the Yandex app, which offers a fluid user-experience across various Yandex services, including localized weather notifications so our users can plan accordingly. Yandex has also integrated Yandex.Weather with their intelligent assistant, Alice. Users can ask their Alice-equipped devices about the forecast, whether they are on the go or at home. Alice shares weather updates to users on their phones, laptops, connected cars, and smart speakers.
Yandex is able to offer nowcasting to a greater number of Yandex.Weather users now that they have access to data from a second satellite, the Himawari-8. Previously, one stationary weather satellite, the Meteosat-8 over East Africa, provided the data for Yandex.Weather’s nowcasting. The Himawari-8’s position over New Guinea provides Meteum with weather data from a larger area that Meteosat-8 doesn’t cover. As a result, Yandex.Weather users in Siberia and the Russian Far East now have hyperlocal nowcasting for the first time. Thanks to the positioning of the satellites, Yandex.Weather can even provide users with nowcasting for a wide variety of locations outside their primary market of Russia, stretching from London to Shanghai. In the future, Yandex.Weather engineers plan to pull data from a diverse range of weather satellites so the service’s nowcasting can be offered to users wherever they may be in the world.
While Yandex users east of the Ural Mountains now have access to minute-by-minute forecasting throughout the Yandex ecosystem, they’ve long been able to use Yandex.Weather for Russia’s most accurate hourly and daily weather predictions. When comparing weather predictions in Russia to the European Centre for Medium-Range Weather Forecasts (ECMWF), Yandex.Weather is up to 42% more accurate in predicting hourly precipitation, and up to 28% more accurate when predicting the 10-day temperature forecast.
Accurate weather forecasts are not only vital to how Yandex's users plan their daily routines but can also influence how they interact with various Yandex services. Yandex will continue to innovate their weather service for users and explore new ways to incorporate Yandex.Weather’s forecasting throughout their ecosystem.