Researchers have developed a new hybrid model merging two conventional ways of measuring air pollutants such as ozone. Air quality predictions have been found to be far more efficient with the hybrid model than any existing models1.
Usually, two approaches are used to predict ambient air quality — the deterministic and the statistical. The deterministic approach accurately predicts concentration of air pollutants in the middle percentile range and the statistical models provide a better estimate of concentrations in the extreme percentile ranges. However, the statistical models are site specific and are unable to generate 'what-if' scenarios while the deterministic models are general in character and useful in creating alternative scenarios.
The researchers merged the benefits of the two models to create a hybrid model that can predict the 'entire range' of distribution. They used the hybrid model to predict reactive secondary pollutants like ground level ozone (GLO).
In testing the hybrid model, they used the results of a community multi-scale air quality model at a major traffic intersection as the deterministic component along with the statistical distribution model to predict the entire range of GLO concentrations. The performance of the model was found to improve from an index of agreement from 0.77 (deterministic model) to 0.91 (hybrid model). The results show an improvement in the predictions using the hybrid model over the deterministic model.
The hybrid model could replace conventional models in air quality management programmes.
