Most incinerators modelled by Plume Plotter don't yet exist, in which case it is impossible to measure their pollution.
Even for operational incinerators, monitoring their pollution would require a grid of accurate and sensitive real-time monitors for each pollutant. This would be extremely expensive, and some pollutants can't be monitored in real time.
Plume Plotter uses AERMOD to predict the incinerator pollution, as explained here. AERMOD is one of the world's most widely used air pollution modelling systems and is very popular in the UK, along with a rival product, ADMS. Both AERMOD and ADMS have been extensively validated (e.g. see here) against real-world air pollution. They are not perfect but are continuously being improved. Users of AERMOD and ADMS are not expected to validate the software themselves.
All developers proposing to build an incinerator. They use AERMOD or ADMS to predict the fallout from the incinerator, given certain assumptions. They never validate their predictions after their incinerator is built. Some examples:
Environment Agencies for England, Wales, Scotland and Northern Ireland (and other countries). These agencies require incinerator developers to use emissions dispersal models. Their predictions are added to existing pollutant levels and compared with statutory limits. The agencies use the results to decide whether to issue a permit for an incinerator or to require alterations before granting one.
Researchers investigating links between air pollution and health. Several research projects have used the same tools to predict incinerator fallout. They use this as an estimate of the population's exposure to pollutants and check for links between this exposure and health problems. Some examples:
For most incinerators that Plume Plotter has modelled, the modelling results done by the developer are also publicly available. This document compares Plume Plotter's prediction of the annual mean fallout of nitrogen dioxide with the developer's prediction, for several incinerators. The predictions are generally very similar, even though the modelling uses different weather data and sometimes different software (AERMOD vs. ADMS). This agreement suggests that both were done correctly.