Datasets:
PV Location vs GSP Location
Hi, many thanks to amazing research paper and your README.md information!
Thank you for the fantastic research paper and the information in your README.md!
I am deeply exploring your research paper and the GitHub code.
However, I encountered some confusion regarding the spatial slicing process.
When we randomly select a location defined by GSP (out of 317 GSPs), we spatially slice all other datasets (NWPs, Satellites, and PV) with respect to its center.
For NWP and Satellite data, this makes sense in terms of their window dimensions (roi_height and roi_width).
However, how does this work for PV data? Are we summing all PV generation amounts within the chosen window size based on GSP center?
Additionally, is there a relationship between the PV data and GSPs?
Thank you for your time and consideration!
Hi @kwon-encored thank you very much for getting in contact
Are technic was to use PV data based on a window size too, based on GSP center. Really we should also label if this PV is in a GSP, but we didnt get round to that.
Yea, there is a relationship between PV data and GSPs. Sheffield Solar - https://www.solar.sheffield.ac.uk/pvlive/, take the PV data and use an algorithm to estimate the total amount of PV data in that GSP.
Hope that helps, and feel free to ask any more questions