Online Program Home
My Program

Abstract Details

Activity Number: 295
Type: Topic Contributed
Date/Time: Tuesday, August 2, 2016 : 8:30 AM to 10:20 AM
Sponsor: Royal Statistical Society
Abstract #320108 View Presentation
Title: Functional Regression and Model Calibration for Air Pollution Data
Author(s): Marian Scott* and Alan Hills and Lauren sim
Companies: University of Glasgow and Scottish Environment Protection Agency and University of Glasgow
Keywords: functional data ; air quality ; model calibration

Air quality modelling is a crucial tool for assessing and managing air quality. Spatio-temporal atmospheric dispersion models such as ADMS are increasingly used in policy and management evaluation supported by monitoring data. To evaluate how well the modelled and monitoring data are calibrated, methods such as functional principal components analysis (PCA) and clustering, and functional regression have been investigated. This study focuses on the air quality in the city of Aberdeen in 2012. Aberdeen has a total of 220,000 inhabitants and is located on the east coast of Scotland by the North Sea. The monitoring site data are given by the air quality in Scotland website The data are hourly measured pollutant levels at six monitoring sites in Aberdeen, as part of the Automatic Urban and Rural Networks (AURN). The chosen pollutant is nitrogen dioxide (NO2). The ADMS model output are hourly NO2 concentrations provided on a 75mx75m grid. Statistical considerations include the spatial and temporal variation in model and monitored concentrations, spatial coherence of behaviour and the calibration of model and monitored data.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association