Meteoric Smoke Deposition in the Polar Regions: A Comparison of Measurements With Global Atmospheric Models

• Main Authors: Brooke, JSA, Feng, W, Carrillo-Sánchez, JD, Mann, GW, James, AD, Bardeen, CG, Plane, JMC
• Format: Article in Journal/Newspaper
• Language: English
• Published: American Geophysical Union 2017
• Subjects: Antarctic; The Antarctic; Greenland; Antarc*; Antarctica; ice core
• Online Access: https://eprints.whiterose.ac.uk/121209/; https://eprints.whiterose.ac.uk/121209/18/Brooke_et_al-2017-Journal_of_Geophysical_Research__Atmospheres.pdf

The accumulation rate of meteoric smoke particles (MSPs) in ice cores—determined from the trace elements Ir and Pt, and superparamagnetic Fe particles—is significantly higher than expected from the measured vertical fluxes of Na and Fe atoms in the upper mesosphere and the surface deposition of cosmic spherules. The Whole Atmosphere Community Climate Model with the Community Aerosol and Radiation Model for Atmospheres has been used to simulate MSP production, transport, and deposition, using a global MSP input of 7.9 t d⁻¹ based on these other measurements. The modeled MSP deposition rates are smaller than the measurements by factors of ~32 in Greenland and ~12 in Antarctica, even after reanalysis of the Ir/Pt ice core data with inclusion of a volcanic source. Variations of the model deposition scheme and use of the United Kingdom Chemistry and Aerosols model do not improve the agreement. Direct removal of MSP‐nucleated polar stratospheric cloud particles to the surface gives much better agreement, but would result in an unfeasibly high rate of nitrate deposition. The unablated fraction of cosmic dust (~35 t d⁻¹) would provide sufficient Ir and Pt to account for the Antarctic measurements, but the relatively small flux of these large (>3 μm) particles would lead to greater variability in the ice core measurements than is observed, although this would be partly offset if significant fragmentation of cosmic dust particles occurred during atmospheric entry. Future directions to resolve these discrepancies between models and measurements are also discussed.