Inverse Modeling Identifies Efficient Emission Control Strategy for Mitigating PM_2.5 Pollution

The traditional approach to emissions mitigation design (“forward” scenario testing) evaluates candidate emissions-control strategies but does not directly determine the smallest set of emissions reductions sufficient to bring modeled concentrations below a standard.

Here, we invert the planning task and solve a constrained Bayesian inverse problem to estimate the minimum, spatially explicit emissions changes required to meet a fine particulate matter (PM_2.5) concentration target, while determining the distance to compliance across space, precursors, and sectors. We demonstrate the method with a case study of meeting the recently adopted annual primary National Ambient Air Quality Standard across the contiguous United States.