AI Contest

Welcome to the American Meteorological Society 2013-2014 Solar Energy Prediction Contest! This contest is organized by the American Meteorological Society Committees on Artificial Intelligence Applications to Environmental Science, Probability and Statistics, and Earth and Energy. Prizes are sponsored by EarthRisk Technologies, Inc. The contest leaderboard and data are being hosted by Kaggle. Go here to enter.

Motivation

Renewable energy sources, such as solar and wind, offer many environmental advantages over fossil fuels for electricity generation, but the energy produced by them fluctuates with changing weather conditions. Electric utility companies need accurate forecasts of energy production in order to have the right balance of renewable and fossil fuels available. Errors in the forecast could lead to large expenses for the utility from excess fuel consumption or emergency purchases of electricity from neighboring utilities. Power forecasts typically are derived from numerical weather prediction models, but statistical and machine learning techniques are increasingly being used in conjunction with the numerical models to produce more accurate forecasts.

Objective

The goal of this contest is to discover which statistical and machine learning techniques provide the best short term predictions of solar energy production. Contestants will predict the total daily incoming solar energy at 98 Oklahoma Mesonet sites, which will serve as "solar farms" for the contest. Input numerical weather prediction data for the contest comes from the NOAA/ESRL Global Ensemble Forecast System (GEFS) Reforecast Version 2. Data include all 11 ensemble members and the forecast timesteps 12, 15, 18, 21, and 24. Locations of the Mesonet sites relative to the GEFS data are shown in the above figure. Training data will come from 1994-2007. Public testing data will be from 2008-2009. Private testing data for a more recent period will be used for the final evaluation.

Predictions submitted to the contest should be the total daily solar radiation value for each test day at each Mesonet site. The predictions should be submitted in the same CSV format at the training data with the date in the first column and the site predictions in alphabetical order in each subsequent column.

Participants are welcome to use whatever technique they like to optimize the solar energy forecast. You may enter as individuals or as teams. Each participant may only enter the contest under one username. No outside data are allowed, but the creation of derived variables is encouraged.

The top winners of each part of the contest will then be invited to present a talk about their technique and data handling methods at a special session at the 2014 AMS Annual Meeting in February.

Registration and the leaderboard are now live on Kaggle. Submissions are being accepted between now and November 15.  If you have any questions, please contact djgagne@ou.edu.