This paper investigates the possibility of deciding when to apply a (dynamic) traffic management measure to improve the air quality or reduce CO2 emissions, based on a limited set of (measured) data. It is expected that a combination of monitoring and modeling is needed for reliable air quality estimations. Improvements in traffic and emission models are required for high resolution modeling as needed in the dynamic deployment of intelligent transportation system (ITS) measures. The first results of a project set-up around a real-world pilot in the city of Delft (the Netherlands) show that it is possible to measure, simulate, and model traffic, emissions and air quality on individual vehicle level at a signalized intersection. Furthermore, an agent-based traffic signal control optimization is tested in a micro-simulation environment, to demonstrate what can be gained on air quality with more efficient traffic signal control.