Jul 2 2019
During weather forecast, meteorologists use several models and data sources to track the movements and shapes of clouds that may indicate severe storms. Conversely, the use of more expanding weather data sets and looming deadlines almost make it unfeasible for them to track all storm formations—small-scale ones in particular—in real time.
At present, a research team from Penn State, AccuWeather, Inc., and the University of Almería in Spain has developed a computer model to assist forecasters identify possible severe storms more quickly and precisely. The researchers have devised a model based on machine learning linear classifiers—a type of artificial intelligence (AI)—that can detect rotational movements in clouds from the images of satellite that might not have been seen. The new AI solution ran on the Bridges supercomputer at the Pittsburgh Supercomputing Center.
A senior forensic meteorologist at AccuWeather, Steve Wistar, stated that this tool that can focus potentially threatening formations could help in making better forresearchersresearchersresearchersresearchersresearchersecasting.
The very best forecasting incorporates as much data as possible. There’s so much to take in, as the atmosphere is infinitely complex. By using the models and the data we have [in front of us], we’re taking a snapshot of the most complete look of the atmosphere.
Steve Wistar, Senior Forensic Meteorologist, AccuWeather, Inc.
The scientists collaborated with Wistar and other AccuWeather meteorologists to study >50,000 historical U.S. weather satellite pictures. Among these images, experts detected and marked the shape and movement of “comma-shaped” clouds. These cloud patterns are highly related to cyclone formations, potentially leading to severe weather events including thunderstorms, hail, blizzards, and high winds.
Later, the researchers used computer vision and machine learning techniques and developed computers to automatically recognize and identify comma-shaped clouds in satellite images. Subsequently, the computers can help experts by indicating in real time where, in a massive amount of data, could they focus their attention to find the severe weather onset.
Because the comma-shaped cloud is a visual indicator of severe weather events, our scheme can help meteorologists forecast such events.
Rachel Zheng, Study Lead Researcher and Doctoral Student, College of Information Sciences and Technology, Penn State
The scientists discovered that their method has the ability to effectively spot comma-shaped clouds with 99% accuracy, at an average of 40 seconds for each prediction. Also, it can predict 64% of severe weather events, excelling other available severe-weather detection techniques.
“Our method can capture most human-labeled, comma-shaped clouds,” stated Zheng. “Moreover, our method can detect some comma-shaped clouds before they are fully formed, and our detections are sometimes earlier than human eye recognition.”
The calling of our business is to save lives and protect property. The more advanced notice to people that would be affected by a storm, the better we’re providing that service. We’re trying to get the best information out as early as possible.
Steve Wistar, Senior Forensic Meteorologist, AccuWeather, Inc.
This work improves previous work conducted by AccuWeather and a College of IST research team led by professor James Wang, who is the dissertation adviser of Zheng.
James Wang stated that “We recognized when our collaboration began [with AccuWeather in 2010] that a significant challenge facing meteorologists and climatologists was in making sense of the vast and continually increasing amount of data generated by Earth observation satellites, radars and sensor networks. It is essential to have computerized systems analyze and learn from the data so we can provide timely and proper interpretation of the data in time-sensitive applications such as severe-weather forecasting.”
He further stated, “This research is an early attempt to show feasibility of artificial intelligence-based interpretation of weather-related visual information to the research community. More research to integrate this approach with existing numerical weather-prediction models and other simulation models will likely make the weather forecast more accurate and useful to people.”
Wistar concluded that “The benefit [of this research] is calling the attention of a very busy forecaster to something that may have otherwise been overlooked.”
Besides Zheng, Wang, and Wistar, the research team included Yukun Chen, doctoral student in the College of IST; Jianbo Ye, former doctoral student in the College of IST and current applied scientist at Amazon Lab 126; Jia Li, professor of statistics in Penn State’s Eberly College of Science; Jose Piedra-Fernandez, collaborating faculty member at the University of Almería; and Michael Steinberg, senior vice president at AccuWeather, Inc.
The work was partially supported by the National Science Foundation, the Amazon AWS Cloud Credits for Research Program, and the NVIDIA Corporation’s GPU Grant Program, and was reported in the June 6th, 2019, issue of IEEE Transactions on Geoscience and Remote Sensing.