This website uses cookies to make it simpler. Find out more »

Trends

Introduction

Recent decades have seen a great improvement in Northern Ireland's overall air quality. In particular, concentrations of pollutants associated with coal and oil burning, such as sulphur dioxide (SO 2), have decreased substantially.

These pages shows how air quality in Northern Ireland has changed over the years since routine automatic monitoring began in the region. Automatic monitoring of ozone has been routinely carried out in Northern Ireland since 1987 at the rural Lough Navar site. Automatic monitoring of other pollutants (including particulate matter as PM 10, sulphur dioxide and oxides of nitrogen) began in the early 1990s, at urban sites in Belfast. However, until the beginning of the 21st century there were relatively few automatic monitoring sites in the region. During the period 2001 to 2005 in particular, many new monitoring sites were set up, especially in areas where levels of air pollution were expected to be high. This has improved our understanding of Northern Ireland's pollution climate.

Trend analysis has been carried out using Openair: a free, open-source software package of tools for analysis of air pollution data. Openair was initially developed with Natural Environment Research Council (NERC) funding. The Openair project is led by Dr David Carslaw, of Ricardo Energy & Environment and the University of York. Here, the Openair ‘TheilSen’ tool, based on the Theil-Sen statistical method, has been used to determine trends in pollutant concentrations over several years. The trend analysis is based on monthly mean pollutant concentrations, calculated here from hourly mean data. Openair includes an option to ‘de-seasonalise’ the data – i.e. to make statistical modifications to the plotted data to remove the influence of seasonal cycles, thus providing a clearer indication of the overall trend over the relevant time. The ‘de-seasonalise’ option has been used here. When this option is used Openair also fills any gaps in the dataset by a linear interpolation method.

The Openair Theil-Sen trend graphs show the trend as a solid red line, with its 95% confidence intervals as dotted red lines. The trend is given at the top of the graph in green, with confidence intervals shown in square brackets. The trend is given as units (i.e. µg m-3) per year, over the period shown. This may be followed by a symbol, with ‘+’ indicating that the trend is statistically significant at the 0.1 level, ’*’ indicating significance at the 0.05 level, ‘**’ indicating significance at the 0.01 level, and ‘***’ indicating significance at the 0.001 level.