Why Is the Weather So Hard to Predict?

Показать описание

If you think the weather forecast is always wrong, well then we’ve got news for you. In Part 1 of this series about the weather, Julian explains everything you need to know about predicting the forecast and why it’s inherently a chaotic mess of math and hailstorms.

Forecasting the weather is pretty hard stuff, to put it mildly. Meteorologists look at real-time data from numerous sources like weather balloons, buoys, radar, and satellites to make predictions about what might happen next, and even with all that data there’s always some uncertainty. The weather is inherently unpredictable and meteorologists try to provide order for that chaos. In fact, a key principle in chaos theory, the “butterfly effect,” has its origins in weather forecasting.

The term “butterfly effect” was coined in the 60s by meteorologist Edward Lorenz. The idea goes that small changes to initial conditions can have large consequences down the road. The saying you’ll often hear is “a butterfly flapping its wings on one side of the world can cause a hurricane on the other side,” but that’s not how the effect got its name. It actually relates to the resulting shape when Lorenz plotted points to equations representing the motion of a gas on a graph. These plotted points kind of look like a butterfly’s wings and show that small changes can make big differences, but that the outcomes aren’t totally random. Whatever happens still has to be within the realm of possibility.

Today, meteorologists largely rely on two major modeling systems: the American model and the European model. Both run on some of the fastest supercomputers in the world, handle tons of variables for things like temperature and pressure, and gather millions of measurements to help handle the initial conditions of the atmosphere. Between the two of them, most forecasters actually agree that the European system is a tad more accurate, mainly because it has a more powerful supercomputing system and can rely on medium-range forecasts. Those are the forecasts between 3-7 days.

#Weather #Climate #Forecast #Meteorology #Terraforming #Seeker #SeekerPlus

--

Read More:
How Reliable Are Weather Forecasts?
“A seven-day forecast can accurately predict the weather about 80 percent of the time and a five-day forecast can accurately predict the weather approximately 90 percent of the time. However, a 10-day—or longer—forecast is only right about half the time.”

When the Butterfly Effect Took Flight
“On a winter day 50 years ago, Edward Lorenz, SM ‘43, ScD ‘48, a mild-mannered meteorology professor at MIT, entered some numbers into a computer program simulating weather patterns and then left his office to get a cup of coffee while the machine ran. When he returned, he noticed a result that would change the course of science.”

One More Coronavirus Problem: Accurate Weather Forecasts
“The National Weather Service uses more than 250 million measurements from aircraft every year, which are fed into complex weather computer models. As of the end of March, meteorological data provided by U.S. aircraft had dropped by half.”

--
Seeker+ is your home for deep dives, fun facts, rabbit holes, and more. Join host Julian Huguet as he unapologetically nerds out on the oddball history, astounding science and intriguing future around topics that will make you the smartest person at your next trivia night.
--
Seeker empowers the curious to understand the science shaping our world. We tell award-winning stories about the natural forces and groundbreaking innovations that impact our lives, our planet, and our universe.

Рекомендации по теме

Комментарии

Head to Colorado. It can be 75 degrees F in the morning, before your mountain hike and by 1;00 pm you can be covered by sleet!!!
Ive seen snow in Colorado's passes in August & other times no snow until late December or no snow at all, all winter!!!

robertmanella

Storm by George R. Stewart (NYRB Classics).

tectorgorch

Thanks for the contenr, I wish you could show visuals, would have loved to see how the butterfly graph looks, what the formula for snow forecast is etc.

Laroac

You mean they AREN'T throwing darts or checking the Magic 8 Ball?? Whoa!😲

Angie_King_Bens_Grandma

It's been a couple years since my meteorology lectures and I have never done anything with it but I think I got something interesting to add.
Here is a couple more reasons why it's so hard to predict the weather and I'll try to keep it short and simple:

Data density/resolution: Location of weather stations is just one small part. We also watch the clouds and there are multiple layers of clouds and atmosphere to consider. It's not just looking at a 2-dimensional map, we have to consider all the layers of atmosphere on top, too. Most of this stuff is either done by weather balloon (it can only collect point data moving up on the y-axis), lidar (like military missile detection systems, they get used to scan for density of raindrops), buoys at sea and satellite data.

Now you might think that satellites provide a lot more dense data but the reality is that most satellites that we use to watch clouds have resolutions of 4x4 or 2x2 km (and that's on the very high end as of a couple years ago). What that means is that a single pixel on a satellite image corresponds to a landscape of 4 km squared. That's a lot of detail that is just getting lost. Strait of Dover for example might appear as all landmass on one of those. Which you can imagine would skew the equations a lot.

But wait, don't we have satellites that can spot people from space? Yes, now we're getting to the most interesting bit.
Never mind the issue of getting access to such satellite data in the first place, the main reason we are so restricted by resolution is the lack of computing power. Even in 2022 we just don't have computers powerful enough, to process all the data we collect. Going from a 4x4 grid to a 2x2 increases the amount of new data to be processed by an insane amount (remember, we
are looking at 3 cardinal directions). Our computers, even today, simply can't keep up.

So in conclusion, we don't have enough data and we don't have enough processing capacity to make use of all the data we have already.

anondynamic