Hobbies And Interests
Computer Modeling
A number of different organizations maintain computer models designed to predict weather patterns and storm movement. Several of these models, such as those used by the European Center for Medium-Range Weather Forecasting, the National Weather Service's Global Forecast Model and the Geophysical Laboratories Fluid Dynamics model, use data collected from all over the planet to create an integrated forecast for the entire planet at once. Others, such as the National Weather Service's North American Model, focus on specific regions and disregard information outside the targeted area. The larger the area covered by the model, the more power it requires to compute a prediction and the longer the run takes.
The Spaghetti Plot
When forecasters plot the aggregated results of these models, they commonly call the resulting graphic a "spaghetti" plot due to the different paths resembling scattered noodles. Often, many of the models will agree closely for the first day or two of predictions, but after that, they may diverge sharply. This is because it is considerably easier to predict weather in the immediate future than it is to predict conditions further out, due to the greater number of variables and uncertainty the models must consider.
Model Differences
Computer models may disagree about the path of a storm for many reasons. Some models use a larger set of input data points than others do, while others may neglect to consider storm intensity in exchange for a better idea of storm direction. These models are extremely complex mathematical equations, and ones that weight variables differently may end up with vastly different results. Each organization refines its model frequently to increase its accuracy, but over time, some have been more successful than others at predicting storm behavior.
Building the Forecast
When the time comes to produce the official forecast, National Weather Service meteorologists consider all the different models they have available to them on the spaghetti plot. If the majority of the models predict movement in a particular direction, that may be a good sign that the storm will head in that direction. In other cases, there may be too much disagreement for a consensus among the models, and the forecasters must then consider which models have been more accurate at predicting storm behavior. The forecast cone represents an amalgamation of the best estimates of the storm's behavior, and the National Weather Service will update it frequently during the life cycle of a tropical storm or hurricane to provide the best prediction possible.