| This talk presents a Bayesian hierarchical model for fusing spatio-temporal atmospheric observations and computer model forecasts. The Bayesian approach not only improves the biased forecasts but also provides a natural way to quantify the uncertainty. We develop a dynamic model which provide flexible framework for forecasting and assessing the associated uncertainties. We illustrate the methods using an example on daily ozone concentration O3 data observed in the eastern United States and output of a computer model known as the Eta Community Multi-scale Air Quality (CMAQ) model. |
|