A study of atmospheric particulate matter in a city of the central region of Brazil using time-series analysis

Uberlandia is an important city of the Brazil central region. In this work, the Total Suspended Particles (TSP) and particles smaller than 10 μm (PM10) were sampled from 2003 to 2006. The Geostatistical technique was used to evaluate the temporal dependence and the time-series analysis with 2004 and 2005 data was used to predict 2006 values. The selection of the best model was conducted using Akaike's information criterion. The obtained results show that the PM10 and TSP series present trend and seasonal components. The SARIMA (0, 2, 1) (0, 2, 1) was the tested model that presented the best predictions.