Skip to main content

Table 4 Statistical scores of a meta-model approximation trained on \(N_{train} = 3000\) compared to the scores of the model chain using the full air quality model

From: Meta-modeling of a simulation chain for urban air quality

State Estimation \(c({\mathbf {p}})\) vs. \({\hat{c}}({\mathbf {p}})\) (\(N_{train} = 3000\)) Statistical scores over 1-month simulations
Over the full grid MNRMSE (14) \(\sqrt{\text {NMSE}}\) (11) Correlation (12)
Meta-model vs. full model chain 0.35 0.30 0.91
At receptor locations NRMSE (10) \(\sqrt{\text {NMSE}}\) (11) Correlation (12)
Meta-model vs. full model chain 0.30 0.24 0.94