Thu. Jan 20th, 2022


People have tried To predict climate change for millennia, using early science – the “red sky at night” is a hopeful seagull for weather-weary sailors who are actually involved with dry air and high pressure in a region – as well as roofs, hand-held observation-drawn maps, and thumbs. Local rules. These guides for future weather forecasting were based on years of observation and experience.

Then, in 1950, a team of mathematicians, meteorologists and computer scientists – led by John von Neumann, a famous mathematician who assisted in the Manhattan Project a few years ago, and Jules Charney, an atmospheric physicist often considered the father of dynamic meteorology – were the first computerized autonomous. Tested.

Charney, with a team of five meteorologists, has divided the United States (by today’s standards) into fairly large parcels, each with an area of ​​more than 700 kilometers. By running a basic algorithm that captures real-time pressure fields in each isolated unit and predicts it within a day, the team created four 24-hour atmospheric forecasts covering the entire country. It took 33 full days and nights to complete the forecast. Although far from perfect, the results were encouraging enough to set a revolution in weather forecasting, leading the field to computer-based modeling.

In the following decades, billions of dollars of investment and the evolution of faster, smaller computers have increased the ability to make predictions. Models are now able to explain the motility of small atmospheric parcels as small as 3 kilometers, and since 1960 these models have been able to incorporate more-accurate data sent from weather satellites.

In 2016 and 2018, GOES-16 and -17 satellites launched into orbit, providing many improvements, including high-resolution images and pinpoint lightning detection. The most popular numerical models, the US-based Global Forecasting System (GFS) and the European Center for Medium-Range Weather Forecasts (ECMWF), have announced significant upgrades this year, and new products and models are being clipped faster than ever before. At the touch of a finger, we can access a surprisingly precise weather forecast for our exact location on the Earth’s surface.

Today’s lightning speed predictions, advanced algorithms and worldwide data collection products are displayed one step away from complete automation. But they are not yet perfect. Despite having expensive models, advanced satellite arrays, and mega-computers, human predictors have a unique set of tools of their own. Experience — the ability to observe and draw connections where algorithms cannot দেয় gives these predictors an edge that surpasses glossy weather machines in the most hazardous conditions.

Although very useful According to Andrew Devanas, operational forecaster at the National Weather Service Office in Key West, Florida, models with large-picture forecasts are not sensitive, say, a slight updraft in a small land quadrant that suggests creating a reservoir. The Devanas Water Spout lives close to one of the most active areas in the world for marine-based tornadoes that could damage the Florida Strait # and even ships coming ashore.

The same constraint prevents the forecasting of thunderstorms, extreme rainfall and land-based tornadoes, e.g. Tear up More than 60 people were killed in the Midwest in early December. But when there are tornadoes on the ground, forecasters can often detect them by looking for their signature on the radar; Waterspouts are very small and often lack this signal. In a tropical environment like Florida Kiss, the weather doesn’t change very much day by day, so Devonas and his colleagues had to manually look for variations in the atmosphere, such as wind speed and available humidity – which algorithms do not. Always consider – to see if there is a link between certain factors and the high risk of flooding. They compare these observations with a modeled probability indicator that indicates that water spots are likely and Found That in the right combination of atmospheric measurements, human prediction Each of the waterfall prediction metrics has “excluded” the model

Similarly, Research David Novak, director of the NOAA Weather Prediction Service, and his colleagues have revealed that although human forecasters may not be able to “beat” models on your normal sunny, fair weather day, they make more accurate weather forecasts than bad algorithm-crunchers. The data that Novak’s team studied over two decades showed that people were 20 to 40 percent more accurate in predicting near-future rainfall than the Global Forecast System (GFS) and the North American Mesoscale Forecast System (NAM), the most widely used national model. People have made statistically significant improvements in temperature forecasting as per the guidelines of both models. “Often, we see that at large events when predictors can make some value-addition improvements to the auto-guide,” Novak says.



Source link

By admin

Leave a Reply

Your email address will not be published. Required fields are marked *