Global variability in atmospheric new particle formation mechanisms

A key challenge in aerosol pollution studies and climate change assessment is to understand how atmospheric aerosol particles are initially formed1,2. Although new particle formation (NPF) mechanisms have been described at specific sites3–6, in most regions, such mechanisms remain uncertain to a large extent because of the limited ability of atmospheric models to simulate critical NPF processes1,7. Here we synthesize molecular-level experiments to develop comprehensive representations of 11 NPF mechanisms and the complex chemical transformation of precursor gases in a fully coupled global climate model. Combined simulations and observations show that the dominant NPF mechanisms are distinct worldwide and vary with region and altitude. Previously neglected or underrepresented mechanisms involving organics, amines, iodine oxoacids and HNO3 probably dominate NPF in most regions with high concentrations of aerosols or large aerosol radiative forcing; such regions include oceanic and human-polluted continental boundary layers, as well as the upper troposphere over rainforests and Asian monsoon regions. These underrepresented mechanisms also play notable roles in other areas, such as the upper troposphere of the Pacific and Atlantic oceans. Accordingly, NPF accounts for different fractions (10–80%) of the nuclei on which cloud forms at 0.5% supersaturation over various regions in the lower troposphere. The comprehensive simulation of global NPF mechanisms can help improve estimation and source attribution of the climate effects of aerosols.

Table 1.To make the figures concise, we show only the concentrations of a precursor in the bestcase simulation and in the sensitivity simulations where its concentrations show significant differences from the best case.

Supplementary Table 1 Summary of model scenarios developed in this study.
1

Scenario Description
Best-case A simulation that includes all 11 NPF mechanisms and uses the R2D-VBS to simulate the nucleating organics.This is our comprehensive best-case scenario and used in most analyses in this study.

No_NPF
A simulation that does not consider any NPF.

NPF_Mech4
A simulation that considers only four traditional inorganic nucleation mechanisms, i.e., the neutral and ion-induced H2SO4-H2O mechanisms and H 2 SO 4 -NH 3 -H 2 O mechanisms, which resembles the NPF treatment in commonly used climate models.
NPF_Mech11_constYield A simulation that includes all 11 NPF mechanisms but assumes that pureorganic and organic-H2SO4 nucleation is driven by a fixed fraction of the monoterpene oxidation products, following the treatment of a number of previous modeling studies 1-4 .The specific "fixed fractions" used here followed Gordon et al. 5 .Specifically, organic-H 2 SO 4 nucleation was linked to all oxidation products of monoterpenes; in other words, the "fixed fraction" of monoterpene oxidation products used to drive organic-H2SO4 nucleation was 1.0.Pure-organic nucleation was assumed driven by highly oxygenated organic molecules (HOMs), the molar yields (fixed fraction) of which were assumed to be 1.4% for the reaction of monoterpenes with O3 and 0.6% for the reaction of monoterpenes with OH.
0.67*SO 2 _0.5*DMSThe same as "Best-case" except that the SO 2 and DMS emissions are reduced by a factor of 1.5 and 2, respectively.
1.5*SO 2 _2*DMS The same as "Best-case" except that the SO 2 and DMS emissions are increased by a factor of 1.5 and 2, respectively.

0.33*H2SO4
The same as "Best-case" except that the simulated H2SO4 concentrations are reduced by a factor of 3.

3*H 2 SO 4
The same as "Best-case" except that the simulated H 2 SO 4 concentrations are increased by a factor of 3.

0.33*MT
The same as "Best-case" except that the monoterpene emissions are reduced by a factor of 3.

org-weak-T-dependence
The same as "Best-case" except that a weaker temperature dependence of pure-organic and organic-H 2 SO 4 nucleation rates is used.

organic-H 2 SO 4 _Riccobono
The same as "Best-case" except that the organic-H 2 SO 4 nucleation parameterization is replaced with the one reported in Riccobono et al.The same as "Best-case" except that the NH 3 concentrations are reduced by a factor of 3.

nonuniform-NH3
The same as "Best-case" except that the NH3 concentration is 1 ppb (consistent with observations in the convective outflow hotspots by Höpfner et al. 8 ) in [average NH 3 ]/1 ppb of the area of each model grid and zero in the remaining area of the model grid in the upper troposphere.For the areas with the presence of NH3, we assume that H2SO4 is exhausted by nucleation.

0.5*HIO3
The same as "Best-case" except that the simulated HIO3 concentrations are reduced by a factor of 2.

5*HIO 3
The same as "Best-case" except that the simulated HIO 3 concentrations are increased by a factor of 5.
upper_tau The time scale for the consumption rate of convective available potential energy (denoted by "tau") is set to the upper bound (14,400 s) of the possible range specified by Qian et al. 9 , as compared to the value of 3,600 s in the best-case simulation.

lower_tau
The time scale for the consumption rate of convective available potential energy (denoted by "tau") is set to the lower bound (1,800 s) of the possible range specified by Qian et al. 9 , as compared to the value of 3,600 s in the best-case simulation. upper_dmpdz The fractional mass entrainment rate (denoted by "dmpdz") is set to the upper bound (2.0 × 10 −3 m −1 ) of the possible range specified by Qian et al. 9 , as compared to 0.7 × 10 −3 m −1 in the best-case simulation.

lower_dmpdz
The fractional mass entrainment rate (denoted by "dmpdz") is set to the lower bound (0.1 × 10 −3 m −1 ) of the possible range specified by Qian et al. 9 , as compared to 0.7 × 10 −3 m −1 in the best-case simulation.

NPF_Mech4_scaled
In this scenario, we applied a fixed scaling factor to the NPF rates in NPF_Mech4 (which only includes four traditional nucleation mechanisms involving H2SO4, NH3, and H2O) such that its globally averaged NPF rate matched that of the best-case simulation.Here, the globally averaged NPF rate is defined as the average of the NPF rates across all model grid boxes (both horizontally and vertically), weighted by the volumes of those grid boxes.
1 Supplementary Table 2 Iodine chemical reactions implemented in the model.
a The reaction is predicted to be negligible at the altitudes of interest Supplementary Table 3