i think its only giving a concentration because to model the spatial dispersion pattern of a hypothetical volume of *some contaminant* in air, there has to be a starting concentration of *some contaminant*. it looks like they've used a dust based model, and the standard for quantifying concentrations of stuff in air when the exact composition is unknown (like dust / ash / etc.) is mass per cubic vol. for such hypotheticals the exact values are irrelevant and meaningless, but their relative differences are still useful as a vector.

and I wasn't sure of any of this either at first so i simply clicked the link OP provided to read about the model and its assumptions. There is an impressive amount of documentation available for HYSPLIT. I have learned some cool stuff already. Most of your questions / assumptions could have been addressed by reading the documentation, and doing so would probably even make clear how many of the shortcomings you've identified in the model could be improved. But it doesn't seem like you're interested in learning / contributing / helping bc all you've done so far is use hundreds of words to say nothing more than "youre dumb thats dumb you cant do that", and that seems like a waste of your knowledge