The impact of nitrogen chemistry in snow on atmospheric oxidising capacity in the polar boundary layer

Hoi Chan

Research output: ThesisDoctoral Thesis

152 Downloads (Pure)


Snow photochemical reactions drive production of chemical trace gases within snowpack, including nitrogen oxides (NOx = NO + NO2), which are then released to the lower atmosphere. They were found to affect the oxidis- ing capacity of the lower troposphere, especially in remote regions of high latitudes with low level of anthropogenic pollution, by altering concentra- tion of ozone (O3), a pollutant and green house gas, and hydroxyl radical (OH), responsible for the removal of many atmospheric pollutants including methane. However, the emission of NOx and its impact has not yet been quantified. Current air-snow coupled models are limited by poor under- standing of interactions between air and ice and often require fitted parame- ters to match the model results with observations that limited their predictive capability.
Presented here is a new 1-D air-snow exchange model for nitrogen species to investigate snow as the source of NOx in the overlying atmosphere in Antarctica. Building on existing models, it includes heat and radiation trans- fer in snow, gas and solid phase diffusion, multi-phase chemistry and air- snow grain exchange of oxidised nitrogen species. The solar radiation trans- fer in snow is parameterised as an exponential decay with respect to depth to reduce computation cost. The impact of the exponential parameterisation on the photochemical production of NOx is evaluated by comparing the re- sults with the outputs from a radiative transfer model - TUV (Lee-Taylor and Madronich, 2002). The study showed the exponential parameterisation of radiation of cold polar snowpack has no significant impacts on the estima- tion of photochemical production rate due to large solar zenith angles and the efficient light scattering environment of the snowpack. For other types of snowpack, such as melting or fresh snow, the overestimation in photochem- ical production at large solar zenith angles or underestimation at small solar zenith angles caused by the exponential parameterisation can be corrected by applying the solar zenith angle and chemical species dependent correc- tion factor.
Two temperature dependent multi-phase air-snow grain exchange mod- els were developed from physically based parameterisations, each based on different hypothesis on the interface between air and snow grain. The first model assumed at temperatures below a threshold temperature, To, the air- snow grain interface is pure ice and above To, a disordered interface (DI) emerges covering the entire surface of the snow grain, which is a similar ap- proach taken by models previously developed. The other model assumed at temperatures below melting of ice, the air-snow grain interface is pure ice and liquid assumed to be co-existed with ice as micropocket at temperatures above the eutectic temperature. They are validated with existing Antarctic snow samples and atmospheric observations from a cold site on the Antarctica Plateau (Dome C, 75◦ 06′ S, 123◦ 33′ E, 3233 m a.s.l.) and at a relatively warm site on the Antarctica coast (Halley, 75◦35′S,26◦39′E, 35 m a.s.l). The study showed the concentration of nitrate in surface snow is better described by the latter model, that it reproduced a good agreement with observations at both sites without requiring any tuning parameters whereas the first model only showed good agreement at Dome C but not at Halley. It is therefore suggested that in winter the air-snow exchange of nitrate is determined by non-equilibrium adsorption on ice and co-condensation coupled with diffusion to the bulk ice. In summer, the air-snow exchange of nitrate is dominated by solvation into liquid micropocket following Henry’s law.
The 1-D air-snow model was developed based on an existing model framework and the newly developed temperature dependent multi-phase air-snow grain exchange models to estimate the flux of NO3 – to the overlying atmosphere and the distribution of NOx within the snowpack. The model predicted an average emission flux of NO2, FNO2 , is 3 × 1012 molecule m−2 s−1 at Dome C, in late December.
Original languageEnglish
Awarding Institution
  • Royal Holloway, University of London
  • King, Martin, Supervisor
  • Frey, Markus, Supervisor, External person
Thesis sponsors
Award date1 Feb 2018
Publication statusUnpublished - 17 Jan 2018


  • air-snow model
  • photochemistry
  • Polar Region

Cite this