Attribution of future US ozone pollution to regional emissions, climate change, long-range transport, and model deficiency

Abstract. A regional chemical transport model (CTM) is used to quantify the relative contributions of future US ozone pollution from regional emissions, climate change, long-range transport (LRT) of pollutants, and model deficiency. After incorporating dynamic lateral boundary conditions (LBCs) from a global CTM, the representation of present-day US ozone pollution is notably improved. This nested system projects substantial surface ozone trends for 2050's: 6–10 ppbv decreases under the "clean" A1B scenario and ~15 ppbv increases under the "dirty" A1Fi scenario. Among the total trends, regional emissions changes dominate, contributing negative 20–50% in A1B and positive 20–40% in A1Fi, while LRT effects through chemical LBCs and climate changes account for respectively 15–50% and 10–30% in both scenarios. The projection uncertainty due to model biases is region dependent, ranging from −10 to 50%. It is shown that model biases of present-day simulations can propagate into future projections systematically but nonlinearly, and the accurate specification of LBCs is essential for US ozone projections.