Table of contents:
- How does the weather arise
- How natural phenomena were studied before
- Up-to-date weather forecasts
- Are perfect predictions possible?
Forecasters promise a sunny day, and outside the window - a blizzard. Inaccuracies in forecasts are associated with both rapidly changing environmental conditions and global climate transformations. Nevertheless, modern meteorologists have made a breakthrough in forecasting, today mathematical algorithms are used for this, new methods and tools are being created to study current weather conditions.
How natural phenomena are being studied today and whether it is possible to make perfectly accurate forecasts in the future was readily available in the American popular science magazine Discover. T&P adapted the article and translated it.
How does the weather arise
The Earth's atmospheric layer is composed primarily of nitrogen, oxygen, and water vapor. This air behaves like a liquid and, flowing from one place to another, changes temperature, humidity and other characteristics. Weather is a byproduct of the atmosphere, which carries heat from one place to another.
Cooler air is dense and cannot hold much moisture; warmer air is less dense and can hold more water. When areas of air of different temperature and density collide, rain occurs as water is formed from the cooled warm air. Other precipitation may occur. As the warm, humid air rises upward, it cools and the water condenses on particles suspended in the air. Rising droplets become heavier and subsequently fall to the Earth.
A hurricane is formed when seawater heats above 27 ℃ and evaporates rapidly, and the air above the ocean becomes warm and rises. In its place, streams of cold air come, which also heat up and rise. These movements create strong winds, a hurricane is formed.
How natural phenomena were studied before
Scientific weather observations began in the Renaissance, when barometers and thermometers were invented. Ancient European scholars such as Galileo used these tools to explain weather phenomena.
But early predictions were limited and based only on the assumption that the past would determine future behavior.
“If there’s a storm today in Kansas and the next day in Missouri, you can tell it will come to Illinois the next day,” explains Bob Henson, meteorologist and author of Weather Underground.
This method works under constant conditions - when the storm is gradually moving or when the local climate does not change much from day to day (for example, in Southern California).
However, this simple method does not take into account changing conditions: for example, storms are quickly formed due to convection (movement of air volumes from one altitude to another, due to the Archimedean force. - Ed.). Fortunately, there are new ways to predict the weather in today's world. Forecasts are not made by people looking at the maps and yesterday's highs and lows, they are made by machines.
Up-to-date weather forecasts
Meteorologists use numerical weather forecasting by entering data about current weather conditions. Then they are processed in a computer model. The more relevant and accurate information is entered, the more accurate the forecast will be. To obtain this data, instruments such as a weather balloon, airplanes, satellites, and ocean buoys are used.
Weather patterns divide a region, state, or even the entire globe into cells. Their size affects the accuracy of the forecast.Large rectangles are harder to see what is happening in small areas, but they provide an overall picture of weather trends over time. This general forecast is necessary, for example, in order to determine the movement of a storm.
Small cells with higher resolution allow forecasting over a shorter time period - one to two days - and only cover a specific area. Some models may focus on specific data such as wind speed, temperature and humidity. Therefore, two computer models can give slightly different results even with exactly the same initial observations.
Are perfect predictions possible?
“Computer models are sufficient for day-to-day weather forecasts, so meteorologists will not add much here,” says Schumacher, assistant professor of atmospheric sciences at Colorado State University. - But this does not mean that people are not needed at all. The forecaster can recognize inaccuracies in the data reproduced by the computer system."
Precipitation is much harder to predict than temperature, says Matt Kölsch, a hydrometeorologist at the University's Atmospheric Research Corporation in Boulder.
“Temperature is a continuous field, it is everywhere. Precipitation is an intermittent field, in some places it may rain or snow, but in others it will not be at all."
The landscape features of the area, such as hills, coastlines, affect precipitation, and computer models do not always take this into account. Kölsch says a meteorologist is needed to make forecasts for 24 to 36 hours. Predicting high impact situations such as hurricanes, tornadoes and floods is more challenging and requires both human resources and computer systems.
Since the 1950s, fast computers have become more and more accurate predictions. Today's five-day forecast is accurate about 90% of the time, according to the National Oceanic and Atmospheric Administration.
The 7-day forecast turns out to be correct 80% of the time, and the 10-day forecast 50%
Today, a five-day hurricane forecast is more reliable than a four-day forecast in the early 2000s and a three-day forecast in the 1990s. And a 2015 Nature paper found that forecasts for three to ten days improved by about a day in a decade, which means that the current six-day forecast is as accurate as the five-day forecast 10 years ago.
Unfortunately, major climatic transformations complicate the forecasting process. There is a joke that a butterfly flapping in Hong Kong can change the weather in New York. This idea was put forward in 1972 by the mathematician and meteorologist Edward Lorenz. The “butterfly effect” is that small changes can have a huge impact on the development of the entire system.
In practice, this means that one computer model, run more than once, even with the smallest differences in the current conditions, can give different predictions. Because of this, the potential forecasting limit is around 14 days, says Bob Henson.
“Lorenz basically said that it’s impossible to predict the weather patterns for more than two weeks, because these little butterfly wings and countless other little things will lead to big changes,” says Henson.
Meteorologist Judt is confident that humanity can never predict a thunderstorm more than a couple of hours in advance, no matter how good the observations are.
“For hurricanes and storms that are much stronger (and therefore easier to detect in advance), the period can be two to three weeks,” he says.
When making a forecast, meteorologists account for uncertainties by using a mathematical model several times. At the same time, it will give a slightly different result, but most of them will be similar.The most frequent ones will be the final result.