Interpretation of concentration fields in 3D particle mode

[aseagram]

I am wondering how to interpret the concentration output when using 3D particle mode (INITD = 0), especially with respect to computing a vertical smoke (column) integration like the product NOAA provides (http://airquality.weather.gov/sectors/conus.php).

To my knowledge, when using 3D particle mode, the concentration fields can be interpreted as the total concentration (particles) between the level and the level below. For example, if the levels were output at 0 m and 5000 m, then the concentration field should be the total concentration between 0 and 5000 m.

I am wondering what the effect is if setting many vertical levels and then summing the grids to obtain a vertical integration. i.e.:
(1) VERTICAL_LEVELS = 0 5000 , where the total column concentration is the output at 5000 m
versus
(2) VERTICAL_LEVELS = 0 100 250 500 1000 2000 4000 5000 , where the total column concentration is obtained by summing the concentrations at all the levels.

Should these two options result in the same concentration? If not, why?
Are the resulting total fields to be interpreted in the same way? If not, what do they physically mean?

I have run an initial test case and find that there are very large differences between using two vertical levels (option 1) and using many and then summing them (option 2).

Thanks for your advice.

[ariel.stein]

These two options will not give you the same result since the vertical distribution of the lagrangian particles is not uniform with height up to 5000 m. In order to calculated the concentrations, the model divides the mass of the lagrangian particles by the volume in which the particles reside. The concentrations calculated within a well mixed boundary layer (e.g up to 2000 m) should be independent of the choice of vertical layers. However, above the boundary layer the distribution is not uniform producing some layers to be populated with particles and other not.

On the other hand, the column integration (concentration x height) should give you equivalent results. If you multiply the concentration obtained with Option (1) by 5000 you get the column integrated pollutant. On the other hand if you multiply each concentration segment in option (2) by its corresponding height and add them up then you should get the same column integrated quantity.