Modelling Long-Term Pattern Pollutants Using Recurrence Quantification Tools. Case-study: PM10 at Mexico City

Abstract Various techniques have been used to model the long-term trends for particulate matter (PM) in the environment. However, the non-linear behavior of airborne pollution (PM10, in this case) and the variety of compounds present in the environment make long-term pattern modelling a challenging task. In this contribution, Recurrence plots and its extension RQA (Recurrence Quantification Analysis) explore the long-term trends on PM10 in Mexico City. To determine the feasibility of this technique over a long period of time, data was obtained for various monitoring stations from different stations over 12 years. The trends resulted were then compared with statistical analyses from other authors. The results confirm that this technique could be used to model long-term patterns of particulate matter given that the right RQA tools are used. Furthermore, the importance of this work relies on the fact that this type of analysis with such an amount of data using RPs has not been carried out before, which makes this tool to find PM10 trends buried in large databases, useful for this case analysis. Also, different approaches have been performed and preliminary conclusions are drawn from these experiments.