Is beam focusing accurate to 3D beams in a 2D simulation?

[plasmahomie]

Hi,
I was testing the simulation of a focusing Gaussian laser beam in 2D EPOCH as described in the docs (https://epochpic.github.io/quickstart/basic_examples/laser.html) and found that the focus position and beam radius are correctly reproduced but the intensity at focus is underestimated. This is the same result as in the docs, and the underestimated intensity is explained as due to the paraxial approximation not being entirely applicable there (1.7 um spot, 1.0 um wavelength). However I was simulating a 5um spot with wavelength 0.8um and saw similar underestimation. Even with a 30 um spot the underestimation persisted.

So 3D focusing is not accurately reproduced in 2D EPOCH. I think this is because 2D EPOCH can only simulate the convergence of phase fronts in one plane. To reach the correct higher intensity energy needs to come into the central slice from the z axis, and this cannot be reproduced in a 2D simulation. Is this a known phenomenon? If so I think the docs should really be modified to acknowledge this, as myself and others I know have definitely misplaced confidence in 2D simulations, assuming the focusing is correctly simulated as the docs suggest.