4. Results and discussion
4.1. Additivity test
The additivity test has revealed that for NO2, the modelling
hypotheses implemented during the construction of AERIS seem
reasonable. The main question here was assessing whether a TM on
its own is able to reproduce a change in the pollutant concentration
in the same way that this change is produced by the AQMS along
with other varying sectors. The goal of the analysis is to gain insight
on the flexibility of the modelling framework of AERIS.
Fig. 4 depicts the behaviour of the different cells of the domain
for different thresholds of the additivity test. It can be seen that
most cells (4416e98.1%) are below the lowest threshold, covering
most of the modelling domain while a minority of cells are above
t ¼ 1.0 mg/m3 and t ¼ 1.5 mg/m3 (73 and 11, 1.6% and 0.2% respectively).
The spatial representation of this behaviour is relevant to
detect the presence of problematic zones (i.e. hotspots) and to
assure that the estimates are valid throughout the modelled
domain. Domain-averaged results for the additivity test are presented
in Table 8 for each of the evaluated sectors. As it can be seen,
the resulting NO2 mean annual concentrations calculated by both
models are similar, with low mean absolute errors (jXi,AERISXi,AQMSj<0.5
mg/m3
). Additionally, the normalized mean bias (NMB)
for the complete domain equalled 4.1% and the normalized mean
error 13.0% (NME), being both values below the NMB15% and
NME35% limits for acceptable performance according to Boylan
and Russell (2006). It should be noted that the estimates
การแปล กรุณารอสักครู่..
