![]() ![]() The machine learning program correctly highlighted the corridor where widespread heavy rain and flooding would occur. The right panel shows the resulting observations of excessive rainfall. 31, more than 24 hours ahead of the event. The left panel shows the forecast probability of excessive rainfall, available on the morning of Aug. An excessive rainfall forecast from the Colorado State University-Machine Learning Probabilities system for the extreme rainfall associated with the remnants of Hurricane Ida in the mid-Atlantic states in September 2021. National Weather Service forecasters are using some of these tools to better assess the likelihood of hazardous weather on a given day. Other research groups are developing similar tools. We have since applied similar methods to forecasting of tornadoes, large hail and severe thunderstorm winds. The result is a tool that forecasts the probability that rains heavy enough to generate flash flooding will occur. We plugged that information into a machine learning method known as “ random forests,” which uses many decision trees to split a mass of data and predict the likelihood of different outcomes. We accounted for this by using observations of unusually large rain accumulations for each location across the country, along with a history of forecasts from a numerical weather prediction model. ![]() ![]() Of course, part of the problem is defining “heavy”: Two inches of rain in New Orleans may mean something very different than in Phoenix. In a project that started in 2017 and was reported in a 2021 paper, we focused on heavy rainfall. Forecasters are using these tools in several ways now, including making predictions of high-impact weather that the models can’t provide. Furthermore, experienced forecasters are remarkably good at synthesizing the huge amounts of weather information they have to consider each day, but their memories and bandwidth are not infinite.Īrtificial intelligence and machine learning can help with some of these challenges. As an example, think of a thunderstorm that dumps heavy rain on one side of town and nothing on the other side. These models are outstanding at predicting most weather systems, but the smaller a weather event is, the more difficult it is to predict. These models use observations of the current state of the atmosphere from sources such as weather stations, weather balloons and satellites, and solve equations that govern the motion of air. Today, weather forecasters’ primary tools are numerical weather prediction models. Australian meteorologist Dean Narramore explains why it’s hard to forecast large thunderstorms. 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. We have seen some of what’s possible in our research on applying machine learning to forecasts of high-impact weather. Chaos limits our ability to make precise forecasts beyond about 10 days.Īs in many other scientific fields, the proliferation of tools like artificial intelligence and machine learning holds great promise for weather prediction. And then there’s chaos, often described as the “butterfly effect” – the fact that small changes in complex processes make weather less predictable. Thunderstorms that produce tornadoes, large hail or heavy rain remain difficult to predict. Errors in the predicted tracks of hurricanes have been cut in half in the last 30 years. Especially in recent decades, steady progress in research, data and computing has enabled a “ quiet revolution of numerical weather prediction.”įor example, a forecast of heavy rainfall two days in advance is now as good as a same-day forecast was in the mid-1990s. But that is a dream,” Richardson concluded.Ī century later, modern weather forecasts are based on the kind of complex computations that Richardson imagined – and they’ve become more accurate than anything he envisioned. “Perhaps 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. It didn’t work because not enough was known about the science of the atmosphere at that time. In his 1922 book, “ Weather Prediction By Numerical Process,” Richardson tried to write an equation that he could use to solve the dynamics of the atmosphere based on hand calculations. Russ Schumacher, Colorado State University and Aaron Hill, Colorado State UniversityĪ century ago, English mathematician Lewis Fry Richardson proposed a startling idea for that time: constructing a systematic process based on math for predicting the weather. Hyoung Chang/The Denver Post via Getty Images ![]() Meteorologist Todd Dankers monitors weather patterns in Boulder, Colorado, Oct. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |