*Christopher Muscato*Show bio

Chris has a master's degree in history and teaches at the University of Northern Colorado.

Lesson Transcript

Instructor:
*Christopher Muscato*
Show bio

Chris has a master's degree in history and teaches at the University of Northern Colorado.

The weather can be forecasted using many different methods. Learn how to identify and describe the types of weather forecasting: persistence forecasting, synoptic forecasting, statistical forecasting, and computer forecasting.
Updated: 11/12/2021

What will the weather be like tomorrow? If you could answer that with absolute certainty, not only would you have eliminated 90% of small talk in this country, but you would also be extraordinarily rich. Whether in anticipation of a vacation or hopeful day off of school, or to gauge the success of crops for a season, people are always trying to predict the weather. Being able to do so with absolute certainty would be paramount to telling the future, so it's impossible, but we do have some good ways to at least make an educated guess.

**Weather forecasting** is the scientific prediction of the state of atmospheric conditions, or a hypothesis about the weather based on reliable data. It seems that people have always wondered about the weather, and in fact, we have found attempted weather forecasts for as long as we have a written record of human history. So, let's check out some of the techniques we've come up with, and just maybe we'll be able to predict the weather.

Let's start with the basics. If today is 72 degrees and mostly sunny with slight rain showers in the afternoon, I'm assuming that it's not going to be 10 degrees and snowing tomorrow. That's the general idea behind our first method, called **persistence forecasting**, or the prediction of future weather based on the assumption that current weather trends will not change. In short, the weather trends will persist. This is the simplest method of forecasting and can be applied to short-term forecasting, as in tomorrow will probably look like today, or long-term forecasting - for example predicting that a hot, dry month of summer will be followed by another hot and dry month. Persistence forecasting works well in places like southern California where weather is very consistent. Where I live in Colorado, weather is much more erratic, so this method is not as successful.

Sometimes you do want a slightly more sophisticated method of forecasting, so how about **synoptic forecasting**, predicting weather based on large-scale meteorological patterns? You see, if something is synoptic, it is a general summary, a view of several parts at a common point. In this case, that means a common point in time. Basically, when you make your forecast, you look at weather patterns on a large scale, observing various pressure systems, weather fronts, and other events. Now, when I say a large scale, I mean 1,000 kilometers or roughly 620 miles. That's the **synoptic scale**. If you've ever seen these sorts of maps on the nightly news, that's synoptic forecasting in action. The idea is that weather patterns affect each other in predictable ways, so by looking at the larger system we can predict changes in the near future. For example, this cyclone in the Pacific is generating various weather patterns that we can predict will change the pressure over us, leading to heavier than usual rain.

Let's take this one step further. Thanks to all of our modern science that can record various aspects of the weather, we can compare various sets of data to predict the likelihood of various future weather patterns. That's the idea behind **statistical forecasting**, the use of statistical models to objectively predict the weather. There are many ways to do this, but the most basic is **linear regression**, or the linear relationship between variables. One variable changes, and the other responds because they are related. Wind increases, temperature decreases. Now, that's a simple model, and this can get much more complex as more variables are added. In forecasting, the independent variables that can suggest weather patterns are called **predictors**. These can be anything from past weather to current air pressure and climatological data, but it's always important to select appropriate predictors for your forecast, otherwise the results could be way off.

People who use statistical forecasting like it because it removes human subjectivity, at least theoretically. It's also nice because, as a statistical model, it calculates and accounts for degrees of uncertainty. The lower the degree of uncertainty, the greater statistic probability that your forecast will be correct.

Now, this is the 21st century, so there's no point in talking about forecasting without computer assistance. Computer-based forecasting is called **numerical weather prediction**, or NWP. This uses computer programs called **forecast models** to account for as many variables as the computer can hold, from past weather trends to current temperature, air pressure, precipitation, etc. The forecast model virtually imitates atmospheric behavior to predict how these variables will realistically interact. In other words, the program is designed to act like real weather.

Now, frankly, a computer program cannot always behave like a natural environmental system, so there will always be some level of error in these predictions, but this is what most modern meteorologists use. NWP forecasting is pretty reliable for up to about ten days, which is why most news channels have a 10-day forecast, but beyond that, the model starts to break down. So, it's not perfect, but it's the latest in a tradition dating back millennia, as humans keep trying to predict tomorrow's weather.

**Weather forecasting** is the scientific prediction of the state of atmospheric conditions, and it's something humans have been at for pretty much as long as we were able. Today, we have a few techniques we rely on. **Persistence forecasting** is the prediction of future weather based on the assumption that current weather trends will not change. This is the simplest method and generally accurate in areas with consistent weather. For a more sophisticated method, try **synoptic forecasting**, predicting weather based on large-scale meteorological patterns. This uses the **synoptic scale** of 1000 km or roughly 620 miles to predict major changes. We can also use **statistical forecasting**, or the use of statistical models to objectively predict the weather. Some people like this for its subjectivity and accounting for degrees of uncertainty. Most modern meteorologists will also use computer-based forecasting called **numerical weather prediction**, to virtually imitate atmospheric behavior to predict how variables will realistically interact. None of these is perfect, but together they can be used to give us a pretty good glimpse into our meteorological future.

To unlock this lesson you must be a Study.com Member.

Create your account

Are you a student or a teacher?

Already a member? Log In

BackRelated Study Materials

Browse by subject