Humans have tried To anticipate the vagaries of climate for thousands of years, using early lore – a “red sky at night” is an optimistic nod to weather-weary sailors that actually correlates with dry air and high pressure over an area – as well as observations from rooftops, hand-drawn maps and basic local rules. This evidence for future weather predictions was based on years of observation and experience.
Then, in the 1950s, a group of mathematicians, meteorologists, and computer scientists — led by John von Neumann, the famous mathematician who had helped with the Manhattan Project years earlier, and Jules Charney, an atmospheric physicist often considered the father of dynamic meteorology I tested the first computerized automated prediction.
Charney, with a team of five meteorologists, divided the United States (by today’s standards) into relatively large pieces, each over 700 kilometers in size. By running a basic algorithm that took the real-time pressure field in each separate unit and projected it forward throughout the day, the team created four 24-hour weather forecasts covering the entire country. It took 33 days and a full night to complete the forecast. Although the results are far from perfect, they were encouraging enough to revolutionize weather forecasting, and move the field toward computer-based modeling.
Over the following decades, billions of dollars in investments and the development of faster and smaller computers led to an increase in predictive power. Models are now able to explain the dynamics of atmospheric chunks of up to 3 km in size, and since 1960 these models have been able to include more accurate data sent from weather satellites.
In 2016 and 2018, the GOES-16 and -17 satellites were launched into orbit, providing a range of improvements, including high-resolution imagery and lightning identification. The most popular digital models, the US-based Global Forecasting System (GFS) and the European Center for Medium-Range Weather Forecasts (ECMWF), have released significant upgrades this year, and new products and models are being developed at a clip faster than ever before. With the touch of a finger, we can access amazingly accurate weather forecasts for our exact location on the Earth’s surface.
Today’s lightning-fast forecasts, the product of advanced algorithms and global data collection, are one step closer to full automation. But it’s not perfect yet. Despite expensive models, an advanced array of satellites, and huge computers, human forecasters have their own unique set of tools. Experience—their ability to observe and plot connections where algorithms can’t—gives these forecasters an edge that continues to outperform flashy weather machines in the most dangerous situations.
Although it is very useful With the big-picture prediction, the models aren’t sensitive to, for example, a small upwelling in one small quarter of the land indicating a stream is forming, according to Andrew Devanas, an operational forecast expert at the Office of the National Weather Service in Key West, Florida. Devanas live near one of the world’s most active areas for waterpipes, sea hurricanes that can damage ships passing through the #Florida Strait and even ashore.
The same limitation hampers predictions of thunderstorms, heavy rain, and tornadoes, such as the one that swept across the Midwest in early December, killing more than 60 people. But when hurricanes occur on Earth, forecasters can often detect them by looking for their signature on radar; Water pipes are much smaller and often lack this signal. In a tropical environment like the Florida Keys, the weather doesn’t change much from day to day, so Devanas and his colleagues had to manually look at differences in the atmosphere, such as wind speed and available humidity — differences that algorithms don’t always keep in mind — to see if there were Any correlation between certain factors and the increased risk of water pipes occurrence. They compared these observations to a typical probability index indicating whether water mains are likely and found that with the right combination of atmospheric measurements, the human prediction “outperformed” the model on every waterpipe prediction measure.
Similarly, research published by NOAA Weather Forecasting Service Director David Novak and colleagues shows that while human forecasters may not be able to “beat” models on your typical sunny day and mild weather, they can still provide more accurate predictions than algorithm-grinding machines in bad weather. Over the two decades of information Novak’s team studied, humans have been 20 to 40 percent more accurate in forecasting near-future precipitation than the Global Forecasting System (GFS) and the North American Medium Range Forecasting System (NAM), the most widely used national models. . Humans also made statistically significant improvements to temperature prediction over the guidelines of both models. “Often we find that it’s in larger events that is when forecasters can make some value-added improvement over automated guidance,” Novak says.
[ad_2]
Source link
