The microphysical properties of tropical convective anvil cirrus: A comparison of models and observations

P. R.A. Brown, A. J. Heymsfield

Research output: Contribution to journalArticlepeer-review

22 Scopus citations

Abstract

This paper describes the use of two-dimensional (2-D) cloud-resolving numerical model simulations of a period of active convection from the Tropical Ocean-Global Atmosphere Coupled Ocean-Atmosphere Response Experiment to generate statistics of the distribution of total ice water content (IWC) and of the relative contribution to IWC of small (sub-200 μm) particles with temperature. The model data are sampled in a way which excludes contributions from active convective updraught regions, so that the model results may be compared with similar such distributions derived from in situ sampling of a number of tropical anvil ice clouds in the vicinity of Kwajalein in the west Pacific. The model gives a good description of the mean and mode IWC in each of seven temperature ranges and, for temperatures warmer than-40°C, correctly predicts the dominant contribution of large particles to the IWC. At lower temperatures, the model retains an excessive fraction of its IWC in large particles. Estimates of the maximum crystal length (MCL) that would be sampled by a 2-D Optical Array Probe exceed the observed values in this temperature range but are otherwise in good agreement. Tests of the sensitivity of model results to the cloud ice bulk density, and the absence of a homogeneous freezing source of ice crystals, suggest that the excessive MCL values are due in part to excessive rates of autoconversion and aggregation of cloud ice.

Original languageEnglish
Pages (from-to)1535-1550
Number of pages16
JournalQuarterly Journal of the Royal Meteorological Society
Volume127
Issue number575
DOIs
StatePublished - Jul 2001

Keywords

  • Anvil cloud
  • Cloud-resolving model
  • Microphysics

Fingerprint

Dive into the research topics of 'The microphysical properties of tropical convective anvil cirrus: A comparison of models and observations'. Together they form a unique fingerprint.

Cite this