Man vs. Machine, Round 2

Some of you may remember that 2 years ago, I did a man vs. machine forecast, where I made a forecast without any model data, and then put the model forecasts up to see who would be closer. I've been interested in doing that again, so I'll be doing it again this week. The only big change, this time the contest will last Monday through Friday (meaning a forecast everyday this week). The forecast site is Alton, IL. Why? I am here, and I can make physical observations if needed.

Just like last time, the only data I will have access to is surface data, upper air data, radar, satellite, and the Skew-T ln-P chart. With the Skew-T, most of them have computed values on the chart. I will be ignoring those, and focusing only on the chart itself. If I want to look at values from the chart, I must calculate them myself. After I write my forecast, I will then look at the model data to put up their forecast, but just to put up their forecast. For numerical values (such as temps) I will use the model's MOS guidance for KALN.

For those who didn't see this contest last time, here's why I do it. A couple of years ago, I was in a rush to finish a forecast, and I pulled model data to finish, even though I didn't agree with it. It was calling for overcast skies, and I didn't think it would happen. I figured I was a young forecaster and the model is a supercomputer cranking out equations, it is probably right. Next day, not a cloud in the sky. After that bust, I said I would only ever use models for forecast guidance, but never again would I trust the model without looking at everything. So, I did the man vs. machine, and I think I won, since all my numbers were closer to the actual results than the computers.

I figure it is time to do this again, and this time the challenge is increased. Instead of just one day, this will be a week. I'm looking forward to this, and I hope you are too. This is going to be fun!