- Lightning is one of nature’s most destructive forces, as seen by a big fire in California’s Lightning Complex in 2020, but it’s also one of the hardest to anticipate. According to a recent study funded by the University of Washington, machine learning, a computer system that improves itself without human programming, can be used to enhance lightning prediction.
Better lightning predictions can aid in the prevention of wildfires, the improvement of lightning safety alerts, and the development of more accurate long-term climate models. Weather forecasts are combined with machine learning algorithms based on previous lightning incidents in this innovative approach.
“This demonstrates that machine learning-based technologies may enhance forecasting of hazardous weather systems like thunderstorms.” Cheng, Wei-Yi He worked on his atmospheric science Ph.D. at the University of Washington. “It stimulates the use of machine learning approaches for forecasting other forms of stormy weather, such as tornadoes and hailstorms.”
Researchers used lightning data from 2010 to 2016 to train the system, which allowed computers to figure out the link between climatic factors and lightning strikes. From 2017 to 2019, we put our method to the test in the weather, comparing AI-assisted ways to existing physics-based techniques and comparing the two using real-world lightning data.
In lightning-rich locations like the southern United States, the new approach anticipated lightning with the same accuracy as the primary method roughly two days earlier. Because it had taught this approach across the United States, its performance in locations where lightning was less prevalent was less precise.
The method employed in the comparison was a newly developed method for forecasting lightning based on precipitation and storm cloud ascent rate. The technique predicts an increase in lightning due to climate change and growth in lightning in the Arctic.
“The current technique multiplies only two variables. It is simple and stems from human thought. However, predicting lightning using these two characteristics isn’t always the best option. “No,” Kim responds.
“Machine learning demands a lot of data, which is one of the requirements that machine learning algorithms must meet to do useful tasks,” Kim explained. In the United States, there is currently a commercial network of equipment for lightning monitoring. A new geostationary satellite continually watches one area from orbit and gives precise lightning data, allowing for more excellent machine learning.
“The amount and quality of the data are the most important factors,” Cheng explains, “and this is what WWLLN can supply us.” “As machine learning technology progresses, having precise and trustworthy lightning observation datasets becomes increasingly vital.”
Researchers hope to enhance the process with new data sources, meteorological conditions, and sophisticated technology. They seek to improve their forecasting for specific scenarios like dry thunderstorms and lightning without rain. These are particularly hazardous in the event of a wildfire.
The researchers believe it may also use their technology to make longer-distance predictions. Because lightning has an impact on atmospheric chemistry, long-distance patterns are crucial. As a result, forecasting lightning leads to a more accurate climate model. Yes, I am. “I was one of the skeptics, but the findings of this and other research have persuaded me.”
Auxiliary collaborators Scott Henderson Robert Holzworth Yugunham, and Jeonghwan Kim were both students at Chonnam National University in South Korea when they met at the University of Washington.
The world’s first optical oscilloscope is created:
A team at the University of Central Florida created the first optical oscilloscope, which can read the electric field of light.
An oscilloscope is a device that visually shows electrical signals and illustrates how power, intensity, frequency, and other parameters vary over time. Light oscillations are converted into electrical impulses displayed on an oscilloscope. The increasing frequency of light waves allows them to communicate much larger information, making interpreting their electric field more challenging.
It can only resolve the average signal produced by a light pulse with current technology, not the peaks and valleys.
The signal’s peaks and troughs are essential since this is where data may be stored and conveyed. The technology, developed by a team at the University of Central Florida (UCF), translates light oscillations into electrical impulses the same way hospital monitors do with a patient’s heartbeat. According to academics, interpreting the electric field of light has been challenging until recently due to the quick speeds at which light waves vary.
Electric fields of up to gigahertz frequencies, encompassing the radio frequency and microwave sections of the electromagnetic spectrum, can now be clocked using the most contemporary technology utilized to power our phone and internet services. Light waves fluctuate at significantly faster rates than sound waves, allowing for a more significant data density to be communicated.
Experts believe it’s critical to measure the peaks and troughs inside a single pulse since that’s where information may be packed and transmitted. “With our optical oscilloscope, we could boost that speed by a factor of 10,000.”
Until now, you’ve had to rely on power. The rapid speed of field light vibrates with this wave of light. Electric fields may now be clocked at maximum gigahertz frequencies, encompassing the radio and microwave areas of the electromagnetic spectrum, thanks to cutting-edge technology that drives modern telephone and internet connections.
“Light communication through optical fiber” To make things quicker, but the speed of the oscilloscope still limits us, “says Michael Chini, a senior associate scientist who collaborated on research at UCF. 10,000. The team’s next goal is to see how fast they can push the technique’s speed limit.
Yangyang Liu, a postdoctoral scholar at UCF, is the paper’s primary author. Jonathan Nesper ’19 ’21 MS, a physics graduate, is one of the other authors. Chini managed the research team and came up with a single-shot waveform measuring system. Beetar assisted in determining the carrier envelope’s phase dependency. All of the writers did the data analysis and published journal publications.