Here are some animations illustrating the spatio-temporal relationships between power at multiple wind farms (top) and wind speed/direction (bottom). Capturing these relationships can improve short-term wind power forecasts.
Second place in EEM20 Wind Power Forecasting Competition! Read all about it here.
ProbCast Released! A beta version of the R package I have been developing with Ciaran Gilbert for probabilistic forecasting is now public on github. You can read about it in our PMAPS paper too.
Slides form Bristol Seminar: for most of February I'm a Heilbronn Visitor in Data Science at the University of Bristol and gave this presentation in the Bristol Data Science Seminar Series.
New Papers recently published on forecast evaluation and containment of loss events on the GB power system. Available here.
New Paper! Ciaran Gilbert recently published his work on improving wind farm power forecasts by leveraging data from individual turbines! Read it here.