{"id":4324,"date":"2022-06-15T22:00:00","date_gmt":"2022-06-15T22:00:00","guid":{"rendered":"https:\/\/modernsciences.org\/staging\/4414\/?p=4324"},"modified":"2022-06-01T08:31:28","modified_gmt":"2022-06-01T08:31:28","slug":"ai-and-machine-learning-are-improving-weather-forecasts-but-they-wont-replace-human-experts","status":"publish","type":"post","link":"https:\/\/modernsciences.org\/staging\/4414\/ai-and-machine-learning-are-improving-weather-forecasts-but-they-wont-replace-human-experts\/","title":{"rendered":"AI and machine learning are improving weather forecasts, but they won\u2019t replace human experts"},"content":{"rendered":"\n
\n \n
\n Meteorologist Todd Dankers monitors weather patterns in Boulder, Colorado, Oct. 24, 2018.\n Hyoung Chang\/The Denver Post via Getty Images<\/a><\/span>\n <\/figcaption>\n <\/figure>\n\nRuss Schumacher<\/a>, Colorado State University<\/a><\/em> and Aaron Hill<\/a>, Colorado State University<\/a><\/em><\/span>\n\n

A century ago, English mathematician Lewis Fry Richardson<\/a> proposed a startling idea for that time: constructing a systematic process based on math for predicting the weather. In his 1922 book, \u201cWeather Prediction By Numerical Process<\/a>,\u201d Richardson tried to write an equation that he could use to solve the dynamics of the atmosphere based on hand calculations.<\/p>\n\n

It didn\u2019t work because not enough was known about the science of the atmosphere at that time. \u201cPerhaps some day in the dim future it will be possible to advance the computations faster than the weather advances and at a cost less than the saving to mankind due to the information gained. But that is a dream,\u201d Richardson concluded. <\/p>\n\n

A century later, modern weather forecasts are based on the kind of complex computations that Richardson imagined<\/a> \u2013 and they\u2019ve become more accurate than anything he envisioned. Especially in recent decades, steady progress in research, data and computing has enabled a \u201cquiet revolution of numerical weather prediction<\/a>.\u201d<\/p>\n\n

For example, a forecast of heavy rainfall two days in advance is now as good<\/a> as a same-day forecast was in the mid-1990s. Errors in the predicted tracks of hurricanes have been cut in half<\/a> in the last 30 years. <\/p>\n\n

There still are major challenges. Thunderstorms that produce tornadoes, large hail or heavy rain remain difficult to predict. And then there\u2019s chaos, often described as the \u201cbutterfly effect\u201d \u2013 the fact that small changes in complex processes make weather less predictable<\/a>. Chaos limits our ability to make precise forecasts beyond about 10 days<\/a>. <\/p>\n\n

As in many other scientific fields, the proliferation of tools like artificial intelligence and machine learning holds great promise for weather prediction. We have seen some of what\u2019s possible in our research<\/a> on applying machine learning<\/a> to forecasts of high-impact weather. But we also believe that while these tools open up new possibilities for better forecasts, many parts of the job are handled more skillfully by experienced people. <\/p>\n\n

\n