Given that the correlation is pretty good, I wonder if it's sufficient to just add a compensating factor to get a better sense of air quality.