WILMINGTON, NC (WECT) - We often talk about the models, "The models say this...," "the models are trending toward that..." But what are they? How do they work? How do we know which ones are right and which ones are wrong? And most importantly why has the most recent run of model data given us hope that the forecast track of Irene will move east?
WHAT IS A MODEL:
The first thing to understand is what exactly is a model. Essentially, a model is a computer program that produces meteorological information for future times at given locations and altitudes. Basically a solution to a forecast using real data from plot points, put into a super computer to determine via complicated fluid dynamic and thermodynamic equations how a forecast will play out. Models are used to compute everything from hurricanes to cold fronts to rainfall totals.
HOW DOES IT WORK:
The initial data that goes into the equation is very important. If a storm is moving over a remote section of the Atlantic the initial data might be scarce. That means the final resolution of the model will be skewed from the actual outcome. That's why the National hurricane Center uses the hurricane Hunter Aircrafts to "investigate" a storm dropping instruments to closely analyze conditions in the storm. This data is then used to fill in the blanks for some of the various forecast models.
Good data is the most important factor as extremely small errors in the initial input, such as temperatures and winds, within numerical models can double every five days, leaving forecaster scratching their heads.
Each model has different parameters and equations that it "favors" or is geared to determine certain aspects of forecasting atmospheric conditions. It is important to mention that there is not a perfect model. Most are good at handling one or two particular aspects of the forecast. The other thing to mention is some are set up to favor historical behaviors while others still examine and focus on current conditions to come up with a forecast.
WHAT DOES IT TELL US:
Once the models are rendered, which happens at predetermined regularly scheduled intervals throughout the day, they spit out a solution. Some of these can be in minute or hourly intervals forecasting as long as the next month, or as short as the next six hours. While each model may vary widely in the forecast it comes up with there are models that, well, use other models as a way to come up with a forecast.
Using such consensus forecast models, as well as ensemble members of the various models, helps reduce forecast error. However, regardless how small the average error becomes with any individual system, large errors within any particular piece of guidance are still possible on any given model run.
HOW DO WE USE THEM?:
Simply put, as guides. Knowing what a model says will happen is important, but knowing why a model comes up with that particular solution is even MORE important. Another important factor is how the models relate to one another. Do they all have different outcomes, or are they similar in their solution? When they are similar, that is what we refer to as "high confidence". Models landing a hurricane or cold front in a similar location give forecasters a higher degree of certainty to how an event will play out.
WHAT IT ALL MEANS:
For the general public model data can frequently be misunderstood. Trying to differentiate which model does what can be a struggle. That's why with all of this talk of math and models you still need the human element. Meteorologists are required to interpret the model data into weather forecasts that are understandable to the viewer. We can also use knowledge of local effects which may be too small in size to be resolved by the model to add information to the forecast.