how is a weather forecast made?

Question

how is a 3 months weather forecast made?
methods of conducting a weather forecast

Answer 1

Data is fed into a computer of atmospheric pressure, temperature, humidity, wind speed, etc gathered by weather balloons, satellites and doppler radar and a pattern is formed on the computer screen that meteorologists can read and make a prediction that is not very reliable for long range forecasts.

Answer 2

I am sort of amused and sort of insulted by several of the previous answers.

A lot of work goes into preparing a forecast.

But, what I find interesting is that you can prepare a forecast by each of the methods mentioned. A forecast is nothing more than an opinion of future events based on that person's knowledge, experience, and practice.

In some cases, maybe the dartboard method would prove to be correct. But, in most cases, I think a practicing forecaster would win.

Anyone who has ever, walked outside, looked up, and said, "I think it is going to rain." is practicing weather forecasting whether they are a meteorologist or not.

If you really want to know what goes into a forecast, you should visit a working forecast office. Most offices will accommodate someone who is really interested in knowing and allow a person to sit along side a forecaster for half a day or more to observe what it takes to put a forecast together.

To try to describe everything would be inappropriate for this forum.

Answer 3

They probably go by each specific day from years past. They usually only can give a temp high and a low for something that far in advance. The weather channel can give you a 10 day forecast, but you can only really trust the next 48 hours, if that.

Answer 4

I am answering this not so much with the hope of it being chosen as best but to inform any interested readers of a couple misconceptions which are common here. If a few of them come away with a better understanding of this after reading my answer, I accomplished something. I am amazed at how often people answering these questions think the 2 things below are an important aspect of producing a forecast: 1 barometric pressure readings to construct maps which forecasts are based on and pressure changes 2 comparing the current conditions with past conditions or only those of which are similar to determine what the future weather will be based on what happened then Regarding 1, while it is certainly true that the postion and movement of Highs and Lows on a weather map provide alot of useful information, atmospheric pressure is only one of the *many* variables being measured and predicted by the computer models which most forecasts are based on. Winds, temperatures, humidity, and "ground conditions" such as vegetation, ocean temperatures, and snow cover all contribute very significantly. The models basically take physical equations and parameterizations for the atmosphere and solve them. Because of the enormous number of calculations needed - issues such as grid spacing and time step required and numerical instability because of those - some of the largest computers in the world are used to make forecasts using data over the entire globe. No human can come close to doing such calculations. Yet one thing humans *can* do is recognize weather systems and use their knowledge of how they develop and progress to forecast. This in some cases allows the human to guess better than the model and identify ways the model may be wrong or unrealistic. That brings us to 2. Looking at past weather events and what happened to forecast current ones is not nearly practical. Even if you used a computer to sort all of them using some sort of criteria of what is similar, there are too many small ways the patterns would differ which can lead to very large differences as the weather progresses. Some people like to use the term chaos regarding that - though there are certain things which the physical situation determines. E.g., you can talk about butterflies flapping their wings until you are blue in the face, but during the Indian monsoon there will be onshore winds and rain in the summer no matter where and how the butterfly flapped its wings 2 weeks ago. One way this is sort of done though is that after a computer model such as the NAM has been run for several years, statistical equations are developed from all past model runs and how they verified to make a numerical forecast. This is called the MOS technique, and basically performs a regression using variables from the model output such as 850 mb temperature at 18 hours, surface pressure at 24 hours, mean 850-500 mb relative humidity at 12 hours, etc. which explain most of the variance. I.e., the variables are found and equations made which when used over all those cases gives the most accurate forecast on the average. --- So to answer your question - as the person above stated, techniques of forecasters can differ. Below is what I tend to do: I look the current conditions - sky cover, temperature, humidity, etc. and data at other locations on the surface map. Look at radar and satellite images. Some other data can be helpful for specific situations such as soundings from radiosondes or satellites. Then I look at model forecasts - which may indeed be the thing I base my forecast on to the largest extent. As I said above, no person can solve all the equations in their head like the model can. I look at the statistical guidance and may even look at the NWS forecast or their discussions to see if they are seeing something I'm not. No forecaster should ever assume they are above that - I can learn something from a 10 year-old. I try to decide to what extent the computer model solution and numerical guidance may be unrealistic based on their biases and what I know and have exprienced about the weather systems involved. Today is a good example - the model guidance forecasted cloudy for us all day, but I felt we would not be seeing a whole lot of clouds until much later when the front nears our region (Poconos). The models forecast the movement of the front well, but too much cloudiness ahead of it - especially the NAM. Yet because of that I thought it would be a bit warmer that it probably will be - still a bit early to say. The model forecast maximum of 79 will probably be better than my 83 despite having the cloud cover wrong. It is about 79 ATTM but still mainly sunny (doesn't typically get a whole lot warmer after midday here). I probably have a warm bias but am usually very accurate despite that.

Answer 5

I`d say they can predict the next week, due to satellites, looking at high and low pressures and where bad weather is situated and the directions Etc... three months ahead would be just by going on historical records....

Answer 6

I am not sure this is possible at all. Most weather forecasters take the money & take the risk of guessing !!!

Answer 7

It all depends on cows and if they are stood up or lying down simple really. You could also consult a pine cone which is probably more accurate than the MET Office.

Answer 8

i laymen terms they look at a sataliet picture of the world. see which clouds, pressures, and winds are about and perdict what will happen in a minute and then from that prediction they predict what will happen the minute after. and they just keep doing that, obviously though the further ahead they predict the least accurate it will become.

hope that helps.

Answer 9

By very clever people using clever equipment, though personally they'd be more accurate if they put their heads out of the window!

Answer 10

It has got to do with high and low pressure and what it brings cloud and wind speed etc.