Cox, M G; Harris, P M; Milton, M J T; Woods, P T (2002) Method for evaluating trends in ozone concentration data and its application to data from the UK Rural Ozone Monitoring Network. NPL Report. CMSC 15/02

Text cmsc15.pdf Download (451kB)

Abstract

A method is developed to evaluate long-term trends in the concentration of ozone at the sixteen sites in the UK Rural Ozone Monitoring Network. The method is based on a regression model incorporating “seasonal” factors that is fitted to the ozone concentration data formed into weekly, monthly, quarterly or yearly averages. The uncertainties associated with the results derived from the regression model are estimated by bootstrap resampling from the residual deviations. This approach has the advantage over traditional approaches that it does not require any assumption to be made about the distribution of the residual errors that are unexplained by the model.
The validity of the method is verified in a number of ways. It is shown that trends estimated from measured data can be sensitive to “missing” data points, and a recommendation is made for using the model in conjunction with data averaged over an appropriate time interval in order to reduce this sensitivity and to have other advantageous properties. Additionally, bootstrap resampling is used to estimate the uncertainty associated with measures of the trend aggregated over all sixteen sites in the UK Rural Ozone Monitoring Network

Item Type:	Report/Guide (NPL Report)
NPL Report No.:	CMSC 15/02
Keywords:	air quality, statistics
Subjects:	Mathematics and Scientific Computing Mathematics and Scientific Computing > Modelling
Last Modified:	02 Feb 2018 13:17
URI:	http://eprintspublications.npl.co.uk/id/eprint/2518

Actions (login required)

View Item