Rights statement: Copyright 1997 by the American Geophysical Union
Final published version, 1.79 MB, PDF document
Final published version
Research output: Contribution to Journal/Magazine › Journal article › peer-review
<mark>Journal publication date</mark> | 20/03/1997 |
---|---|
<mark>Journal</mark> | Journal of Geophysical Research: Atmospheres |
Issue number | D5 |
Volume | 102 |
Number of pages | 18 |
Pages (from-to) | 5953-5970 |
Publication Status | Published |
<mark>Original language</mark> | English |
Simulations of Rn-222 and other short-lived tracers are used to evaluate and intercompare the representations of convective and synoptic processes in 20 global atmospheric transport models. Results show that most established three-dimensional models simulate vertical mixing in the troposphere to within the constraints offered by the observed mean Rn-222 concentrations and that subgrid parameterization of convection is essential for this purpose. However, none of the models captures the observed variability of Rn-222 concentrations in the upper troposphere, and none reproduces the high Rn-222 concentrations measured at 200 hPa over Hawaii. The established three-dimensional models reproduce the frequency and magnitude of high- Rn-222 episodes observed at Crozet Island in the Indian Ocean, demonstrating that they can resolve the synoptic-scale transport of continental plumes with no significant numerical diffusion. Large differences between models are found in the rates of meridional transport in the upper troposphere (interhemispheric exchange, exchange between tropics and high latitudes). The four two-dimensional models which participated in the intercomparison tend to underestimate the rate of vertical transport from the lower to the upper troposphere but show concentrations of Rn-222 in the lower troposphere that are comparable to the zonal mean values in the three-dimensional models.