Solar panel analysis pt 3: Scanning for objects
https://www.jeroenboeye.com/blog/solar-panel-analysis-pt-3-scanning-for-objects/ [www.jeroenboeye.com]
2018-01-03 01:06
While I know that the trees are behind this pattern I was wondering if we could derive their location and shape from the data. If we could somehow get a value for the power loss for each timestamp in the dataset we should be able to couple this data to a position of the Sun at that time using the solarpos function from the maptools package. Then we could create a panoramic picture where each pixel is a position of the Sun throughout the year as seen by the panels. The solar positions that consistently yield too little power should then match the trees.
Interesting series.
source: grugq