If I understand the article correctly this is just the charge difference of the two edges divided by the width of the ribbon. Thus it's possible to compute from the Mulliken populations, although the precise definition of Dy (in the article) is not crystal clear.
You could also plot the electrostatic potential across the ribbon, either at different positions Z along the periodic cell, or averaged over Z. The slope of this is the electric field, which perhaps is not constant across the ribbon, but probably not rapidly varying either, at least not in the averaged case.