A tool to predict cloud cover over Dutch solar parks and quantify the impact on solar power generation.
The main idea is to take the latest satellite cloud imagery, and use wind forecasts to determine where the clouds are going. Then, we compare the forecasted cloud coverage to solar irradiance forecasts in locations of solar parks, to determine where there might be a "hit" to solar generation.
Have a look at notebooks/01_entire_workflow.ipynb for a basic idea of where I'm going with this.