Introduction
Throughout METEO 410, the course takes a hands-on approach to weather forecasting by having the class participate in WxChallenge. WxChallenge is a national forecasting competition that runs for ten weeks during the semester and provides an opportunity to put into practice what has been learned throughout the METEO courses by making short term forecasts for a particular city. Every two weeks, the forecast city changes for a total of five cities by the end of the ten week competition. Each forecast consists of predicting the high and low temperatures, total liquid precipitation, and maximum sustained wind speed for a 24 hour forecast period ranging from 06Z to 06Z.
When creating a forecast, my approach is to look for consensus between the ensembles of computer models and to see how they compare with what MOS is predicting. I also make sure to look for trends in the different forecast runs and see what makes the most sense given the big picture. When the models are consistent, this leads to a high level of confidence for my forecast. However, sometimes they can be extremely inconsistent with a large spread between the ensemble members. This leads to a shaky forecast and the inevitable forecast bust.
The forecast bust that I will be discussing occurred in Little Rock, Arkansas on March 2nd, 2010 during the second week of forecasting for the city. For that day, my predicted high temperature ended up being six degrees warmer than the actual high temperature. Considering that every degree off counts against my score in the competition, it was a pretty big bust.
Before I get into the details surrounding my forecast bust, let’s first take a look at the climatology for Little Rock to get a better sense of what is typical for early March, high temperatures in particular. MOS uses climatology for its predictions so it is important to know when there will be a departure from climatology before making a forecast.
<< Back | Top | Next >>
