Heating Distributions from January and July Simulations of NCAR Community Climate Models

View More View Less
  • 1 Space Science and Engineering Center, University of Wisconsin-Madison, Madison, Wisconsin
© Get Permissions
Full access

Abstract

Simulations of the global distribution of heating (the sum of latent, sensible, short and longwave radiation) are presented for January and July using the R15 NCAR Community Climate Model (CCM). The vertical and horizontal distributions of heating predicted by an earlier version of the CCM (CCM0B) are contrasted with those predicted by the current version of the CCM (CCMI) in which substantial revisions were made in the physical parameterizations of convective, radiative, and sensible heating. The results are compared with climatological studies of atmospheric heating and with recent diagnostic analyses of heating during the Global Weather Experiment (GWE).

The dominant heat sources in the CCM simulations of January and July are located over Indonesia-Southeast Asia in broad agreement with the primary feature of the observed Asian monsoon; however, several marked distinctions between the vertically averaged heating distributions for CCM0B and CCM1 occur. During January, centers of maximum heating are located farther south of the equator in CCM1 than in CCM0B. This southward shift in CCM1 is accompanied by strong heating along the South Pacific and South Atlantic convergence zones. These latter features are largely absent in CCM0B. Additionally, CCM1 heating over the monsoon regions of southern Africa and South America is nearly double that found in CCM0B. Similarly, during July, CCM1 heating in the monsoon regions of northern Africa, the western Pacific, and Central America is nearly double that observed in CCM0B.

With fixed boundary conditions (e.g., sea surface temperatures, soil moisture, sea ice extent, and snow cover) in the perpetual simulations, the interannual variability of heating is due entirely to internal model dynamics. The interannual variability of both January and July heating is larger in CCM1 than in CCM0B. Regions of maximum interannual variability in both CCM0B and CCM1 are found in the vicinity of the principal tropical heat source regions. This variability is associated primarily with in situ fluctuations in the intensity of regional heating centers, while geographical displacements appear to be of secondary importance.

Major differences are found between the vertical distributions of heating for CCM0B and CCM1. These stem largely from changes in physical parameterizations, in particular a change in the prescribed critical relative humidity for condensation by moist stable and unstable adiabatic adjustment from 80% in CCM0B to 100% in CCM1, and a replacement of dry convective adjustment-in CCM0B by vertical diffusion of heat and moisture in CCM1. In the tropics, maximum heating occurs in the lower troposphere in CCM0B, while strongest heating occurs in the mid- to upper troposphere in CCM1. In the storm tracks of extratropical latitudes, heating is confined below 800 mb in CCM0B, while heating of appreciable magnitude extends above 500 mb in CCM1. The vertical distribution of heating in CCM1 agrees favorably with diagnosed distributions for the GWE, while the CCM0B heating distribution does not.

Abstract

Simulations of the global distribution of heating (the sum of latent, sensible, short and longwave radiation) are presented for January and July using the R15 NCAR Community Climate Model (CCM). The vertical and horizontal distributions of heating predicted by an earlier version of the CCM (CCM0B) are contrasted with those predicted by the current version of the CCM (CCMI) in which substantial revisions were made in the physical parameterizations of convective, radiative, and sensible heating. The results are compared with climatological studies of atmospheric heating and with recent diagnostic analyses of heating during the Global Weather Experiment (GWE).

The dominant heat sources in the CCM simulations of January and July are located over Indonesia-Southeast Asia in broad agreement with the primary feature of the observed Asian monsoon; however, several marked distinctions between the vertically averaged heating distributions for CCM0B and CCM1 occur. During January, centers of maximum heating are located farther south of the equator in CCM1 than in CCM0B. This southward shift in CCM1 is accompanied by strong heating along the South Pacific and South Atlantic convergence zones. These latter features are largely absent in CCM0B. Additionally, CCM1 heating over the monsoon regions of southern Africa and South America is nearly double that found in CCM0B. Similarly, during July, CCM1 heating in the monsoon regions of northern Africa, the western Pacific, and Central America is nearly double that observed in CCM0B.

With fixed boundary conditions (e.g., sea surface temperatures, soil moisture, sea ice extent, and snow cover) in the perpetual simulations, the interannual variability of heating is due entirely to internal model dynamics. The interannual variability of both January and July heating is larger in CCM1 than in CCM0B. Regions of maximum interannual variability in both CCM0B and CCM1 are found in the vicinity of the principal tropical heat source regions. This variability is associated primarily with in situ fluctuations in the intensity of regional heating centers, while geographical displacements appear to be of secondary importance.

Major differences are found between the vertical distributions of heating for CCM0B and CCM1. These stem largely from changes in physical parameterizations, in particular a change in the prescribed critical relative humidity for condensation by moist stable and unstable adiabatic adjustment from 80% in CCM0B to 100% in CCM1, and a replacement of dry convective adjustment-in CCM0B by vertical diffusion of heat and moisture in CCM1. In the tropics, maximum heating occurs in the lower troposphere in CCM0B, while strongest heating occurs in the mid- to upper troposphere in CCM1. In the storm tracks of extratropical latitudes, heating is confined below 800 mb in CCM0B, while heating of appreciable magnitude extends above 500 mb in CCM1. The vertical distribution of heating in CCM1 agrees favorably with diagnosed distributions for the GWE, while the CCM0B heating distribution does not.

Save