Where measurements do not exist, we use U.S. eastern and western
regional averages of the modeled:observed ratio (dividing line of
90W), as the model reproduces O3 in the western U.S. much more
accurately than in the East. Like our first paper, O3 exposure,