Microphysical modeling of cirrus: 1. Comparison with 1986 FIRE IFO measurements

Authors:
  • EJEric J. Jensen
    Eric J. Jensen
  • OTOwen B. Toon
    Owen B. Toon
  • DWDouglas L. Westphal
    Douglas L. Westphal
  • SKStefan Kinne
    Stefan Kinne
  • AHAndrew J. Heymsfield
    Andrew J. Heymsfield
Publisher website
Google Scholar
Abstract
We have used a one‐dimensional model of cirrus formation to study the development of cirrus clouds during the 1986 First ISCCP (International Satellite Cloud Climatology Project) Regional Experiment (FIRE) intensive field observations (IFO). The cirrus model includes microphysical, dynamical, and radiative processes. Sulfate aerosols, solution drops, ice crystals, and water vapor are all treated as interactive elements in the model. Ice crystal size distributions are fully resolved based on calculations of homogeneous freezing nucleation, growth by water vapor deposition, evaporation, coagulation, and vertical transport. We have focused on the cirrus observed on November 1, 1986. Vertical wind speed for the one‐dimensional simulation is taken from a mesoscale model simulation for the appropriate time period. The mesoscale model simulation suggested that strong upward motions over Wyoming and subsequent horizontal transport of upper level moisture were responsible for the cirrus observed over Wisconsin on this date. We assumed that our one‐dimensional model could be used to represent a vertical column moving from Wyoming to Wisconsin over a period of several hours. Ice crystal nucleation occurs in our model in the 8 to 10‐km region as a result of the strong updrafts (and cooling) early in the simulation. Growth, coagulation, and sedimentation of these ice crystals result in a broad cloud region (5–10 km thick) with an optical depth of 1–2 after a few hours, in agreement with the FIRE measurements. Comparison with aircraft microphysical measurements made over Wisconsin indicates that the simulation generated reasonable ice water content, but the predicted ice number densities are too low, especially for radii less than about 50 μm. Sensitivity tests suggest that better agreement between simulated and observed microphysical properties is achieved if the nucleation rate is higher or stronger vertical mixing (perhaps associated with multidimensional motions) is present.

This publication has 43 references indexed in Scilit: