This post is part 2 of the "Forecasting Weather" series. The last post is here.
Current weather forecasting mostly focus on temperature, wind and precipitation over a seven day period on a regional scale such as the metropolitan area of St. Louis. Predicting weather for a specific geographical area on a short-term scale, remains a highly challenging computing problem. Using my subdivision that nearly escaped the New Year Eve tornado as an example, to predict the incoming tornado and alert the public (my neighbors), the forecast needs to be provided at a granular scale of a mile and a minute.
And it's this type of "hyper-local" and "near-realtime" forecasting that IBM's Deep Thunder system aims to provide.
Initiated by IBM Research scientists as a collaboration with National Weather Service to forecast weather for 1996 Summer Olympics in Atlanta, Deep Thunder evolved over the years into a full-scale Deep Computing project that focused on much more short-term forecasts, predicting everything from where flooding and downed power lines will likely occur to where winds will be too high for utility works up to 84 hours into the future.
 |
| Deep Thunder's forecast for Hurricane Irene, based on its model for New York's weather, predicted two days ahead that it would be reduced to tropical storm status. (Source: IBM) |
Such forecast capability could be valuable to organizations with weather-sensitive needs. For example, in anticipation of high-wind condition, air traffic control and airlines can take precautionary efforts to reroute air traffic and prevent the situation of massive cancellation or stuck passengers on airplanes.
In another application, a power utility company could learn what areas to prepare for outages in the event of a storm. With this foresight, they could reduce downtime by scheduling maintenance workers to fix a line they expect to fail.
The computing platform behind the Deep Thunder has several key components: receiving and processing data, modeling, post-processing analysis, visualization and dissemination. To achieve predictions with accuracy, Deep Thunder combines weather data from the National Oceanic and Atmospheric Administration, NASA, the U.S. Geological Survey, WeatherBug, and ground sensors.
The simulation codes used for weather modeling has existed for decades. Some trace their origins to the 1970s, but have evolved and improved considerably since then. Instead of inventing new weather models, IBM scientists have been adapting, refining and applying existing models using simulation codes. Additionally, they are developing new methods of data visualization, analysis and dissemination, and techniques for improving computational performance and system automation.
 |
| A readout from Deep Thunder's minute-by-minute windspeed forecasts for lower Manhattan on March 13 from the iOS visualization interface. (source: IBM) |
The government of Rio De Janeiro, Brazil, for example, entered into a partnership with IBM in December 2010 to use Deep Thunder in a new weather prediction center designed to help the city adequately prepare people for flash floods, which left over 200 dead earlier in 2010.
IBM's other partnership is with the University of Brunei Darussalam, which is using the technology at a national level for flood forecasting, and as part of a program to predict the impact of climate change on the country's rainforests.
The Deep Thunder group has also been able to dovetail with other
analytics-driven projects such as Smarter Cities. Working with colleagues in the
new IBM Research center in Brazil as well as the IBM India Research Lab, the
team is leading the Rio de Janeiro project to better anticipate flooding, and
predict where mudslides might be triggered by severe storms. Here, highly
targeted weather modeling is only part of the story. Through a new city command
center, weather data can be integrated with other city information systems to
determine how best to respond to such situations, including where and when to
deploy emergency crews, make optimal use of shelters and monitor hospital bed
availability.
With the World Cup coming to Rio in 2014 and Summer Olympics in 2016, the forecast for the business-of-weather approach pioneered by Deep Thunder looks bright.
With the World Cup coming to Rio in 2014 and Summer Olympics in 2016, the forecast for the business-of-weather approach pioneered by Deep Thunder looks bright.
 |
| The Unified Rio Emergency Operation Center, Powered by Deep Thunder Technology for Hyper-local Weather and Flood Forecast |
IBM's other partnership is with the University of Brunei Darussalam, which is using the technology at a national level for flood forecasting, and as part of a program to predict the impact of climate change on the country's rainforests.
Please comment below about this technology. And make a prediction about when Deep Thunder will be available to forecast weather for my, and your subdivision.
A final word: I wrote this blog on the American Airlines flight from Dallas to San Francisco. It was delayed for about 40 minutes, and the pilot apologetically mentioned that it was due to SFO shutting down two runways tonight, due to high wind…
Links
Links
No comments:
Post a Comment