About the weather: hyperlocal forecast

Weather forecasting, as well as in principle forecasting of something that is difficult to obey by mathematical models, is a complex computer task. However, predicting weather conditions narrowed to a specific location, region, city, or square kilometer, as forecasting risks to the population and infrastructure, is a completely different type of task. It was for its solution that IBM created a system of "hyperlocal" forecasting Deep Thunder .

The peculiarity of the “weather for tomorrow” from Deep Thunder is that, in contrast to significant external conditions for ordinary people, the system focuses on short-term forecasts, trying to predict everything from floods to fallen power lines in the 84-hour time horizon.

Deep Thunder has been developed in the laboratories of the company since 1996 and today it is an indispensable tool for any local government, infrastructure company and other organizations whose activities may be somehow dependent on weather conditions. In fact, all conditions have been created for them - unlike the traditional meteorological map, Deep Thunder can lay out the “weather” in pieces, displaying exactly the data that matters.

Already, there was one buyer who was extremely interested in using Deep Thunder to better manage the infrastructure of the whole city: Rio De Janeiro. The authorities want to prepare as best as possible for the 2014 FIFA World Cup and, of course, the 2016 games, which will also be held in the capital of Brazil.


^{Per-minute forecast of wind strength and direction in lower Manhattan on March 13, iOS loy interface}

In a recent demonstration by New York managers, IBM experts showed everything Deep Thunder can do to present data from the NWS ( National Weather Service ), which , in turn, collects them from sensors on the ground, in the air and from space, adding the data into a mathematical model. All this is processed on a supercomputer (the word Deep in the name hints to us) and, as a result, appears to the user in the form of a three-dimensional cartographic grid moving in the future tense. The NOAA ( National Oceanic and Atmospheric Administration ) even supports several weather models processed on a large cluster inside IBM.

Using iOS and Android applications, you can track key weather indicators (as demonstrated in New York) and set up notifications about specific risks that arise in the desired location. You might think that there is nothing so special about this and, in principle, everyone who depends on the weather (this is a lot more people than it seems at first glance) somehow monitors her forecast. The only difficulty is that based on even such simple information as precipitation and temperature, it is sometimes difficult to conclude.

Take snowfalls, which often overload infrastructure and cause cities to burst at the seams. In January last year in New York it was snowing the same way as in February Moscow, which caused roads and airports to get up almost a day, but surprisingly, no damage was done to power lines. All because there was almost no wind - and the snow did not stick to the wires. It is about such details that Deep Thunder is intended to warn after the user sorts everything that is important for him to see.

The strength and direction of the wind is a parameter that does not interest us in everyday life, for others it is a vital factor. So, all types of high-altitude work are exposed to significant danger in strong winds, and not only: “The difference in application conditions will change greatly if we talk about John F. Kennedy Airport and La Guardia (the smaller airport in the Queens area of ​​New York), instead of Manhattan and Central Park, ”explained Lloyd Treinish, leader of the Thomas Watson Laboratory Environmental Science Team.

Using the data created and analyzed inside Deep Thunder, you can better understand how the weather affects our environment as a whole and act accordingly, which can be even more important than just learning something: you can control forest fires and help people protect themselves in time from a tsunami, a storm, a typhoon and something else that suddenly appeared on their doorstep. The weather does not have to complicate our lives, on the contrary, it can serve us in the agricultural sector, which also remains dependent on precipitation and sunlight.

As I mentioned, at first glance it seems that the weather does not affect anything, but in fact, it remains such a difficult forecast factor that we simply obey it. But for our normal existence, such solutions as Deep Thunder are created, which allow us to learn more about what the weather will be like tomorrow.