How do meteoroligists predict the weather?

Answer 1

With forecast models, composed of upper air soundings, air temp, dewpoint humidity.

Answer 2

Answer, part I - not very well.

Answer, part II - they use a model that looks at the weather now and the recent trends in movement of the weather patterns. The models include "rules" about air mass movement and how these air masses interact at their line of contact (called a "front" or "frontal boundary" among other names).

The models treat air masses as semi-coherent entities that move as a unit despite their not being solid. The catch is that more than one model exists, so frequently they run all of the models and take an average. I understand that for hurricanes, at least 20 different models exist.

Anyway, what happens is that they use these models to run a simulation of the moving air masses. Where there are fronts and moisture flows and such, they can predict rain.

Does the model actually predict rain? No. But past experience tells the forecasters that when you combine humidity at X% and air temperature at Y degrees and ground temperature at Z degrees, you get rain about 40% of the time. And that's where they come up with the numbers. By predicting the movement of air masses and looking at historical data for similar situations.

Now - here's why they have such a spotty track record. It is purely statistical in nature. Even if the conditions are "right" for it, you dont always get that rain. Because of chaos theory and a little thing called SDIC - Sensitive Dependence on Initial Conditions. That is, nature represents a finely balanced situation. Even a tiny nudge in the wrong direction at the wrong moment can affect that balance. Which is why so many folks call meteorologists the "weather guessers."

Answer 3

Dear Mizz, this is a good question and here is a short scientific answer. There is a set of time dependent differential equations with variables such as u, v, and w ( the x, y, and z components of the wind velocity), T (temperature), P (pressure), D (density), which are used to build forecast models (simulations of the true atmospheric behavior). The more sophisticated the model, the better the simulation. The model starts at time zero with the latest observations of these parameters (u,v, etc.) and progresses in time to generate new values for these parameters which become the basis for the next forecast cycle.

Once these forecasts are generated the meteorologist must consider them together with all the other observations and knowledge he or she has regarding their forecast area of interest in order to make an official forecast.

Answer 4

They take into account the surface observations(about temperature, wind,humidity, pressure) taken every three hours throughout the world,upper air observations (giving the values of the same elements as above said) taken every six hours,radar picture and satellite pictures and statistical data etc before deciding the weather of a place.

Answer 5

They do very little thinking these days. Computers pretty much do everything for them. I don't think they could muster up enough balls to make a decision of what to wear in the morning.