PNNL

Abstract

Complex aerosol distributions evolve in space and time as a function of emissions, new particle formation, coagulation, condensational growth, chemical transformation, phase changes, turbulent mixing and transport, removal processes, and ambient meteorological conditions. The ability of chemical transport models to represent the multi-scale processes affecting the lifecycle of aerosols depends on their spatial resolution since aerosol properties are assumed to be constant within a grid cell. While decades of work have gone into developing subgrid treatments of clouds in models, characterizing and treating subgrid variability of aerosol processes have received far less attention. Subgrid-scale-dependent processes that affect aerosol populations could have a significant impact on the formation of particles, their growth to cloud condensation nuclei (CCN) sizes, aerosol-cloud interactions, dry deposition and rainout, and hence their burdens, lifetimes, and radiative forcing. To address this issue, we characterize subgrid-scale variability in terms of measured aerosol number, size, composition, and CCN made by repeated aircraft flight paths over the Atmospheric Radiation Measurement (ARM) program’s Southern Great Plains (SGP) site during the Holistic Interactions of Shallow Clouds, Aerosols and Land Ecosystem (HI-SCALE) campaign. Subgrid variability is quantified in terms of both normalized frequency distributions and percentage difference percentiles using grid spacings of 3, 9, 27, and 81 km that represent those typically used by cloud-system resolving models as well as the current and next generation climate models. Even though the SGP site is a rural location, surprisingly large horizontal gradients in aerosol properties were frequently observed. For example, 90% of the 3, 9, and 27 km cell mean organic matter concentrations were within ~46% of the 81-km cell mean around the SGP site, large spatial variability in aerosol number concentrations and size distributions were found during new particle formation events, and consequently 90% of the 3, 9, and 27 km cell mean CCN number concentrations were within ~38% of the 81-km cell mean. The spatial variability varied seasonally for some aerosol properties, with some having larger spatial variability during the spring and others having larger variability during the late summer or spring. While measurements at a single surface site do not reflect the surrounding variability of aerosol properties at a given time, aircraft measurements that are averaged within an 81-km cell were found to be similar to many, but not all, aerosol properties measured at the ground SGP site. This analysis suggests that it is reasonable to directly compare most ground SGP site aerosol measurements with coarse global climate model predictions. In addition, the variability quantified by the aircraft can be used as an uncertainty range comparing the surface point measurements to model predictions that use coarse grid spacings.