Hello
I am trying to model a multi-stage decision-making process in AMPL; where decisions in one stage influence the constraints and objectives of subsequent stages. Specifically, I have a supply chain optimization problem with demand uncertainty. 
In the first stage; I must decide on production levels & in the second stage, I determine distribution strategies based on the realized demand.
I am struggling with how to structure the model to capture this dependency effectively. Should I use separate models for each stage and link them through shared parameters / is there a way to integrate both stages into a single model in AMPL? 
I have checked DOCS - AMPL Okta documentation guide but still need advice .
Additionally; any guidance on handling uncertainty in this context such as using stochastic programming /scenario-based approaches would be greatly appreciated.
Thank you !