Danger of Leaving Weather Forecast to AI


People have tried anticipating climate change for thousands of years, using ancient myths – the “red night sky” is a symbol of the hope of a weary sailor associated with dry air and high altitude — as well as on-roof, hand-drawn maps, and local toe rules. Future weather forecasts were based on historical events and historical events.

Then, in 1950, a team of mathematicians, astronomers, and computer scientists — led by John von Neumann, a renowned mathematician who had assisted with the Manhattan Project years earlier, and Jule Charney, an astronomer — often observed that he was the father of astronomy — he experimented with computer-based predictions.

Charney, a team of five meteorologists, divided the United States into (according to current standards) quite large packages, each covering more than 450 miles[700 km]in area. Using an early method that took real-time pressure on each area and ensured that it would last for one day, the team made four 24-hour weather forecasts around the country. It took 33 days and nights to fulfill those predictions. Although it was not perfect, the results were so encouraging that it led to climate change, which made the project more computer-based.

Over the next several decades, billions of dollars’ worth of business transactions, along with the evolution of small computers, led to the proliferation of forecasting technology. Models are now capable of interpreting the forces of space forces as small as 3 kilometers in area, and since the 1960s these models have always included accurate data that is always sent from seasonal satellites.

In 2016 and 2018, the GOES-16 and -17 satellites were launched in orbit, offering a number of modifications, including clear images and lightning detection. The most popular digital models, the Global Forecasting System (GFS) of the US-based Global Forecasting System (GFS) and the European Center for Medium-Range Weather Forecasts (ECMWF), have released major changes this year, and new ones and colors are being developed faster than ever. With the touch of a finger, we can sense incredible weather patterns on our planet.

Modern high-speed forecasts, advanced technology and data collection systems around the world, are visible from a distance with complete machines. But they are not perfect now. Although expensive, satellites, and computers, human predators have their own tools. Their experience — their ability to see and record links where algorithms cannot — gives predicts the edge of a path that extends beyond the visually impaired machines in extremely difficult situations.

Even more useful with large-scale forecasting, the models are not affected, say, by a slight increase in one small segment indicating water leakage, according to Andrew Devanas, a project forecaster at the National Weather Service office in Key West, Florida. Devanas live near one of the world’s most active regions due to torrential rains, hurricanes that can destroy ships crossing the Florida Straits # even coming ashore.

The same reduction precludes predicting storm surges, heavy rainfall, and high-altitude storms, such as he tore across the Midwest in early December, killing more than 60 people. the discharge water is very low and usually does not have this sign. In tropical areas such as the Florida Keys, the climate does not change much on a daily basis, so Devanas and his colleagues had to manually look at atmospheric variations, such as wind speed and humidity available – a variation that algorithms do not. always check it out – to see if there was any connection between the other items and the higher risk of a water explosion. They compared this to a possible system that shows if water is flowing out he found that it is an appropriate combination of celestial measurements, human predictions “did very well” in any measure of water shoots.

Similarly, research published by NOAA Weather Prediction Service director David Novak and his colleagues show that even human predators cannot “beat” the models of your solar system, the weather, predict more accurately than algorithm-crunchers. weather. In the two decades of the Novak group’s study, people were 20 to 40 percent more accurate in predicting future rainfall than the Global Forecast System (GFS) and the North American Mesoscale Forecast System (NAM), the most widely used species in the world. People also made significant changes in weather forecasting based on both methods. “Most of the time, we find that in major events it is when the forecasters are able to make some changes to the controls,” says Novak.

Leave a Reply

Your email address will not be published.