Different electron charge densities using parabolic and gcs in MPB model

[Sahar Rabet]

Hi everyone,
I have a system of adsorbate on the uncharged metallic surface (Cu). I ran the simulations using the MPB model with "parabolic" correction. When I calculated the planar average of adsorption electron charge density (rho (slab+ads) -rho(slab) - rho(ads)), I saw a really huge backtail on the back of the slab. But when I used the "gcs" correction, this backtail has completely disappeared. I wonder what could be the cause of this huge difference.
I also tried the same environ files with the Pt-CO example from ENVIRON source, I saw  the same difference but with less intensity. I would be thankful for any insight.

p.s. I attached the ENVIRON input files