Posts under tag: prediction
Over the past weekend, the Mid-Atlantic, New England, and parts of Ontario and Quebec was delivered a massive amount of snow — but not as much as was predicted. FiveThirtyEight.com posted a series of articles on that powerful storm as it arrived and departed the area and were left trying to answer why the prediction models weren’t consistent or in some cases correct.
On the approach of the storm, there were four different weather models used by meteorologists to predict that New York City could get as much as eighteen inches of snow when at actually they received no more than ten (9.8″ measured in Central Park). There are multiple reasons for the variance between models like computational power for the models, the frequency and volume of data gathered, and lack of communication on the margin of error of the forecast.
For more information, check out these articles from FiveThirtyEight.com:
Additionally, CASIT’s Research Support Services (RSS) has data visualization capability for UO programs. For more information about their services and offerings, click here.