As I said, the EnergyPlus AFN does not do this. So, if you are tying to estimate the 1-2 C difference that you might find from stratification like this, you’ll probably need to do some CFD, or maybe edit the IDF to use some of these other Rom Airflow models (eg. the Mundt model), or do some hand calcs.
For hand calcs, this PhD thesis put together a simplified formula from several CFD simulations using inputs of internal heat gains vs incoming flow rates:
https://dspace.mit.edu/handle/1721.1/74907
The formulas in section 8.3.2 could probably give you an estimate of the 1-2C differences you’d find from stratification.
But it’s important to keep in mind that this 1-2C difference from stratification is likely an order of magnitude smaller than the radiant temperature differences that you have in an atrium with an all-glass façade like that. So just remember not to mistake the trees for the forest.