POA from Scada Vs. Global Incident in Coll. Plane

[emangahallah]

I am trying to compare the the Plane of Array Irradiance (POA) in w/m^2 i get from scada with the Global Incident in Coll. Plane from the PVSyst output
knowing that the meteo file i used in simulation is generated using the scada data imported as a custom file
so as i imported the POA to the PVSyst i expected that the POA should be identical with Global Incident in Coll. Plane
but there is a differences in months starting from may to november 
attached a sample of the issue
can any one explain why is this behaviour

thank you in advance

[Auriane Canesse]

Your data looks slightly shifted in time, did you define the time shift correctly when first importing your data?
https://www.pvsyst.com/help/meteo-database/notes-on-meteo/meteonote7-time-shift-in-meteo-files.html?h=poa#time-shift-correction-when-importing-data

When importing POA data, PVsyst will first transpose it to GHI and DHI which are required for shadings calculations. POA will be computed again during the simulation. Slight variations are therefore possible between input POA and GlobInc, but not to the extend you are observing.