Scoring

The FSU Local Forecast Contest is run weekly during Fall and Spring semesters for the dates shown on the contest home page. Forecasts for two sites are made each week (Tallahassee is always one of the sites, and the other site for each week is determined at the beginning of the semester). Forecasts of minimum and maximum temperature and precipitation forecasts are made for two sequential 24-hr periods from 06Z on Saturday to 06Z on Monday. The individual elements of the forecast are described below.


Maximum Temperature
The daily maximum temperature is defined as the highest temperature (in degrees F) measured at a station during the 24-hour period beginning at 0600 UTC. Official verification for the day is given at 0600 UTC the following day in the synoptic code information. The temperature error is scored as the square of the difference between the verified and the forecast temperature:
E(Tmax) = (Tfcst-Tobs)2


Minimum Temperature
The daily minimum temperature is defined as the lowest temperature (in degrees F) measured at a station during the 24-hour period beginning at 0600 UTC. Official verification for the day is given at 0600 UTC the following day in the synoptic code information. The temperature error, E(Tmin), is scored in the same manner as is listed above for the maximum temperature.


Probability of Precipitation
The probability of precipitation (PoP) is the probability that at least 0.01" of precipitation (liquid equivalent for frozen precipitation) will be measured at the station using the official National Weather Service observation equipment during the 24-hour period beginning at 0600 UTC. Official verification for the day is given at 0600 UTC the following day in the synoptic code information. Note that for this contest, a trace of precipitation is scored as No precipitation. For the forecast, percentage between 0 and 100 is entered (an integer value). The error is scored as follows:
E(PoP) = (PoPfcst - 100 * Obs)2 / 100
where Obs is either 1 for precip or 0 for no precipitation/trace. Therefore, errors range between 0 and 100 points. Click HERE for the POP numbers and error totals.


Categorical Precipitation
The precipitation category (Cat) is the amount of precipitation (liquid equivalent for frozen precipitation) which is measured at the station using the official National Weather Service observation equipment during the 24-hour period beginning at 0600 UTC. The categories are:
  Category  Precipitation
0:0.00" -- trace
1:0.01" -- 0.05"
2:0.06" -- 0.24"
3:0.25" -- 0.49"
4:0.50" -- 0.99"
5:1.00" -- 1.99"
6:2.00" +
The error points are calculated as double the difference between the forecast category and the verified category squared:
E(Cat) = [2 * (Catfcst - Catobs)]2
Therefore, for 1 precipitation category missed, the error is 4 pts, 2 categories = 16 pts, 3 categories = 36 pts, etc...


Total Score
The weekly error total is scored as the sum of the 4 previously described forecast quantities, or
ETotal = ETmax + ETmin + EPoP + ECat
If an individual misses a forecast period, they receive the highest forecast score achieved by a valid human entry. The maximum individual score for a week is limited to 1600 points.

The forecast totals for each individual are stored and weekly totals are computed. Users' highest scores may be dropped as the semester progresses. Around the 5-week mark, one drop (a user's highest score) is implemented. A second drop (the 2nd highest score) is implemented later in the semester. Each week, the weekend results and the contest overall totals are computed and stored on this website.