Pakamas Teerasuwannajak

Hello, I have a question regarding the global incident in the collector plane loss diagram, where I noticed approximately a 2% difference between Meteonorm and PVGIS (due to the different meteorological databases). The same assumptions were applied in this project. Could you please clarify the possible causes of this difference? Additionally, I would appreciate your recommendation on which meteorological database is most commonly used in practice. Thank you for your support.