IBM Deep Thunder can predict the weather for the entire country, and for a single city block

Predicting the weather is difficult. Previously, weather forecasters were often joked, because the accuracy of forecasts left much to be desired. However, today the accuracy of weather forecasts has increased significantly due to the improvement of computer technology. Analysis of the dynamics of air currents, modeling of the distribution of clouds in a particular region, assessment of the influence of external factors on weather conditions - all this helps to make forecasts more accurate. But still, weather forecasters are mistaken even now, plus weather is predicted for some rather large regions.

The IBM Deep Thunder service can take weather forecasts to a new level. To do this, our company uses a powerful platform for making weather forecasts from the Weather Channel and its computational capabilities. The result of this merger is the Deep Thunder service. This service can provide accurate weather forecasts not only for a particular city, but also for individual urban neighborhoods.

According to our experts, the weather forecast is an issue that must be solved by physics. “How energy is distributed over the surface, what is the dynamics of evaporation, where clouds are most actively formed, and where to expect thunderstorms,” comments Lloyd Treinish, chief research scientist at IBM. Each of the mentioned situations can be described by a set of equations. Of course, the more accurate the weather forecast you want to make, the more equations must be solved in order to anticipate the influence of even not very important factors. But it's worth it.
Previously, our company worked in conjunction with the Weather Channel, but after some time the partners came to understand the need to merge with this company. IBM plans to collect weather data from around the world, using data from thousands of Weather Channel sensors to compile a live forecast in real time. In processing the data, various algorithms will be used to help detect the smallest weather formation factors.
')

How can I use the service?

Deep Thunder is positioned as a service for business. After all, for the same retail weather is very important. If it is hot, supermarkets can increase the order of mineral water and beverages. If snow falls, the stores should increase the order of the shovels, which will be disassembled by customers affected by the thick snow cover. The weather forecast also matters for other industries. For example, insurance. For example, insurance companies can develop a package of assistance to flood victims, if they learn about the increase in rainfall falling in a particular region. The geographical accuracy of the forecast is several hundred meters.

Information of this kind can be useful not only for retail. For example, airlines that know the exact weather conditions of their region for the next couple of days will be able to calculate the amount of fuel needed for the flight or the weather at the destination airport, which may be important for a soft landing. Using information from other weather platforms, Deep Thunder can make predictions for the near future using existing data.

Among the parameters of weather conditions that are predictable are wind speed and direction, temperature, precipitation. A team of scientists and engineers of more than 3000 people helps us to make accurate forecasts.

“This research is a new word in working with cognitive systems,” says Mary Glackin, head of the science and forecast department at The Weather Company.

What else?

We have other services whose task is to predict certain events in nature. This, for example, is a compilation of a model of what is happening under the earth's surface on IBM Sequoia. Together with IBM Watson and IBM Sequoia, scientists have obtained a system that will significantly increase the level of understanding by scientists of the geological processes leading to the occurrence of natural disasters. And the more data will be, the more accurate the system will be able to get scientists.



IBM helps cities prepare for weather disasters with the help of Intelligent Operations Center for Emergency Management. This new platform combines a visualized real-time weather forecast with a small-scale local map, with important urban infrastructure features. Dangerous places service highlights yellow and red, which immediately attracts attention. Natural and human-induced disasters alone in 2014 claimed 7,700 lives and $ 110 billion, according to data from an insurance company in Switzerland.

Correct weather forecasts - one of the most important components of human life. This also applies to domestic issues, and commerce. Provided that such forecasts are made correctly for different regions and different periods, business can get additional benefits, and clients - peace of mind.