Sulfur dioxide trends in Malta : a statistical computing approach

Abstract

A statistical investigation of data related to emissions and measurement of SO2 in the Maltese islands encompassing the period 2004 to 2012 was conducted. The purpose was to investigate whether SO2 levels were driven by the Marsa power station (MPS), which was considered to be the main source of SO2 on the island. In addition, the study sought to establish spatial and temporal trends in the SO2 concentrations measured throughout the islands. Data was obtained from the Malta Environment and Planning Authority (4 fixed monitoring stations and a diffusion tube network) and also from the Enemalte Corporation (emisions of MPS). This was analysed using the Inter Operability and Automated Mapping Project (IntaMap) and GIS for mapping purposes, as well as R and SPSS packages for statistical processing. The results have shown that average yearly emissions from the MPS decreased from approximately 858 g/hr to 780 g/hr between 2009 and 2012. Diffusion tube and monitoring station data have indicated overall decreases in SO2 with certain localised areas showing increases. It was also determined that there were only two occasions when the 350 μg/m3 hour limit of Directive 2008/50/EC was exceeded. All the stations in the monitoring station network registered higher readings when the winds were Northerly or North-Westerly. The Kordin station was found to have the overall highest SO2 readings while Gharb had the lowest. Results suggested that emissions from the MPS had a more localised effect on SO2 levels compared to previous research. However, a 3-predictor statistical ANCOVA analysis determined that while emissions from the MPS were statistically significant in determining the amount of SO2 being measured in the monitoring stations, the results indicated that there were other contributors. These contributors could have included emissions from the Delimara power station emissions and marine vessels. On the other hand, a 2-predictor model using only readings registered with wind originating from the MPS direction showed that MPS emissions were only statistically relevant for Kordin. Hence, it can be concluded Kordin was the most likely area to be affected by MPS emissions while the effect on Msida, Zejtun and Gharb was negligible. The overall findings of the study indicated that, although the MPS was still found to be a contributor of SO2, other sources should now start to be monitored as well. It is recommended that the identification of new sources of SO2 be a focus of future research, including examination of effects of the Delimara power station and marine vessels.