May 2013 Boards forecast contest

M.T. Cranium · 25-04-2013 05:48AM #1

1. Predict the IMT, an average of five selected stations, which has averaged 11.1 C for the 30-yr period just ended in 2010.

2. Predict the highest daily maximum at any reporting (met.ie) station.

3. Predict the lowest daily minimum at any reporting (met.ie) station.

4. Predict the monthly precip as a percentage of normal (for 11 reporting stations in met.ie Monthly Summary)

5. Predict the monthly sunshine as a percentage of normal (for all reporting stations in met.ie Monthly Summary)

6. Bonus question ... Predict which date in May will produce the largest total rainfall in Ireland. This one needs a bit of explanation:

Each day, Met-E publish a table of daily weather statistics at about 20 stations around the country, under "latest weather / yesterday's weather" ... on rare occasions where the rainfall is missing, we can look it up on the same site under "past weather / daily data." The contest bonus, then, will be to predict which of the 31 days in May will produce the highest total in that table. If you choose a wet day that isn't tops, you'll score some points. Here's the scoring protocol ...

10 pts for heaviest rainfall date correct
8 pts for date of heaviest rainfall, one day off
6 pts for date of heaviest rainfall, two days off
4 pts for date etc three days off
2 pts for date etc four days off

then 9 pts for second heaviest date correct, followed by 7,5,3,1 for close dates.

then 8 pts for third, followed by 6,4,2 for near misses.

then 6 pts for fourth, 3 pts for missing by one day

and finally 5 pts for fifth, 2 pts for missing by one day

Also, the scoring must exceed these minimum values ... 8,8,7,7,7,6,6,6,6 etc
Finally, the score is determined from highest single statistic, you can't combine wet days close to your guess.

I will post an ongoing validation, it's something that will help with my research so it's not unwelcome work for me.

Guesses (on the usual form, see post #2) should just be one number for the date selected, and that should be 1 to 31 ... guesses outside May will be invalid no matter what happens on those dates.

Guesses made after the start of penalty time (0300h 2nd) cannot be for the first of May. The same rule applies to all days thereafter, you can only guess days that have not totally expired. Our time penalties are greater than any advantage you could gain from waiting to see.

Hope you find this to be fun and mildly educational. I'm sure I will.

Late penalties -- As usual, 5 pts deducted for entries on 2nd (0300-1500h), 10pts on 2nd (1500-0300h 3rd) etc ... entries can always be made until penalties reach 100, but we recommend 4th as latest entry or you'll likely score net zero.

Entry form next post, or enter directly on this thread.

M.T. Cranium · 25-04-2013 05:55AM

Just a further note to clarify the bonus question, we're looking at the total rainfalls at all stations, not the highest single amount ... I can't give you an exact number but something like an average rainfall of 10-15 mm around the country would probably set this statistic most years, and one or two places could then see 30-50, if that helps you to visualize what we're forecasting. Look at it this way, you can't miss by more than 15 days if you say the 16th.

Now, here's the form, once again artfully supplied by Missing in Action, I mean Sponge Bob.

https://docs.google.com/spreadsheet/viewform?formkey=dElyZTJQeUJzVjdnQ1JBcThYdHpCSEE6MQ#gid=1