We made one-dimensional optical depth averages by averaging a Bifrost 3D model, and added four different microturbulence recipes to this average. Then we compared Ca~II 854.2 nm synthetic line profiles for these average models to the line profile from the 3D model. From this we found that although the equivalent width of the average model is closer to that of the 3D model when a microturbulence is added, the shapes of the lines are not a good fit. Further, we have split the 3D atmospheres into smaller boxes to look more closely at the fit in different areas. We found that when looking at the smallest boxes microturbulence gives the average lines a worse fit, while in the larger boxes the the addition of microturbulence does not change the overall goodness of fit. When looking into the roots of the microturbulence, we found that there seems to be a weak correlation between the high temperature and velocity areas and the areas with a large difference in the line profile, but neither of these parameters is enough to explain the microturbulence on its own. Overall, the main difference between between the spectra from the 3D model and the avereage model comes from the averaging itself, and the addition of microturbulence reduces the difference on a larger scale, but not on a smaller scale. Therefore, we concluded that the microturbulence works as a smoothing parameter to compensate for 3D motions in the average model, but is not a physical parameter.