Publication

A spatiotemporal calibration and resolution refinement model was fitted to calibrate nitrogen dioxide (NO2) concentration estimates from the Community Multiscale Air Quality (CMAQ) model, using two sources of observed data on NO2 that differed in their spatial and temporal resolutions. To refine the spatial resolution of the CMAQ model estimates, we leveraged information using additional local covariates including total traffic volume within 2 km, population density, elevation, and land use characteristics. Predictions from this model greatly improved the bias in the CMAQ estimates, as observed by the much lower mean squared error (MSE) at the NO2 monitor sites. The final model was used to predict the daily concentration of ambient NO2 over the entire state of Connecticut on a grid with pixels of size 300 x 300 m. A comparison of the prediction map with a similar map for the CMAQ estimates showed marked improvement in the spatial resolution. The effect of local covariates was evident in the finer spatial resolution map, where the contribution of traffic on major highways to ambient NO2 concentration stands out. An animation was also provided to show the change in the concentration of ambient NO2 over space and time for 1994 and 1995.