Table 2. Uncertainties in the air pollution modelling for added capacity at Heathrow
What the reports say
What this means in terms of ‘confidence’
ADMS-Airport as the choice of model
Models attempt to represent future events, using a range of assumptions. Their accuracy can be verified by modelling past events and seeing how close the predictions are. However the assumptions can turn out to be wrong, and the choice of model can lead to incorrect results.
“ADMS-Airport gave the best overall agreement with measured data... NOX was predicted with better than 30% uncertainty... Looking at the sites individually for NOx, there are 5 locations for which the model apparently slightly underestimates, and 2 sites where the model apparently overestimates, which overestimates being by (relatively) larger amounts.”
“Uncertainty assessments on measured data for the London Air Quality Network have shown (and LAQN have adopted) a working uncertainty of ±10% for the measurements of NOx, NO2, and O3 concentrations at the EU Limit Value for annual average NO2 concentrations.”
Atkins (2007) Demonstrating confidence in the PSDH air quality work, pp. 4, 18, 23.
The best aircraft emissions model has uncertainty of up to 30%. It leads to more sites being shown with underestimated than overestimated air pollution levels.
The background air quality data has uncertainty of up to 10%.
2002 is the base year for air pollution modelling
The choice of base year is fundamental to the results of the air pollution model. For instance, use of 2003 (which had unusual meteorological conditions and could act as a ‘bad case’ scenario) would lead to much higher predicted future air pollution levels.
“The year 2003 should be avoided in air quality terms for verification. Whist it is real and clearly worst case, it is widely recognised as an unusual conjugation of meteorological circumstances, and this would not help detailed model verification”
The main government report on air quality baselines12 has predicted future NO2 and NOx levels based on both the 2002 and 2003 baselines. It uses 2003 as the base year, and refers to 2002 as the comparison year. It shows:
using 2002 base year
using 2003 base year
using 2004 base year
Total area (km2) exceeding the annual mean objective of 40 μg m-3, London
DfT 2006, ch. 3 Emission Sources, para. 175
Another anomalous year like 2003 could easily lead to much higher NOx emissions than those predicted using the ‘normal’ base year 2002.
12 Grice, s. T. Bush, J. Stedman, K. Vincent, A. Kent, J. Targa and M. Hobson (2006) Baseline Projections of Air Quality in the UK for the 2006 Review of the Air Quality Strategy, AEA Technology,