Documentation of the PMIP models (Bonfils et al. 1998)

PMIP Documentation for GEN1

National Center for Atmospheric Research / Pennsylvania State University: Model GENESIS 1.02A (R15 L12) 1992

PMIP Representative(s)

Dr. Karl E. Taylor, Program for Climate Model Diagnosis and Intercomparison (PCMDI), Lawrence Livermore National Laboratory, Livermore, CA, 94550, USA, Phone: 925 423 3623, Fax: 925 422 7675; email:


Dr. Lisa C. Sloan, Earth Sciences Department, University of California, Santa Cruz, CA, 95064, USA, Phone: 408 459 3693, Fax: 408 459 3074; email:


Dr. Starley L. Thompson, P.O. Box 3000, National Center for Atmospheric ResearchC0 80307, USA, Phone: 303 497 1628; Fax: 303 497 1348; email:


Dr. David Pollard, Earth System Science Center, Pennsylvania State University, University Park, PA 16802, USA, Phone: 814 863 3673; Fax: 814 865-2022; email:

Model Designation

GENESIS 1.02A (R15 L12) 1992

Model Identification for PMIP


PMIP run(s)

0cal, 21cal.

Number of days in each month: 31 28 31 30 31 30 31 31 30 31 30 31

Model Lineage

The global climate model used for PMIP is Version 1.02A of GENESIS (Global ENvironmental and Ecological Simulation of Interactive Systems). It has been developed at NCAR Interdisciplinary Climate Systems Section. The model consists of an AGCM coupled to multilayer models of vegetation, soil or ice land, snow, sea ice, and 50-m slab oceanic layer. The AGCM originated from NCAR Community Climate Model version 1 (CCM1, described in Williamson et al. 1987), but the physics schemes differ significantly from those of CCM1.

Model Documentation

main references:

Pollard, D., and S.L. Thompson, 1992: Users' guide to the GENESIS Global Climate Model Version 1.02. Interdisciplinary Climate Systems Section, National Center for Atmospheric Research, Boulder, Colorado, 58 pp.

Pollard, D., and S.L. Thompson, 1995: Use of a land-surface-transfer scheme (LSX) in a global climate model: the response to doubling stomatal resistance. Glob. Plan. Change, 10, 129-161.

Thompson, S.L., and D. Pollard, 1995: A global climate model (GENESIS) with a land-surface-transfer scheme

(LSX). Part 1: Present climate simulation. J. Climate, 8, 732-761.

Numerical/Computational Properties

Horizontal Representation

Spectral (spherical harmonic basis functions) with transformation to an appropriate nonuniform Gaussian grid for calculation of nonlinear atmospheric quantities. The surface variables (see Ocean, Sea Ice, Snow Cover, Surface Characteristics, Surface Fluxes, and Land Surface Processes) are computed on a uniform latitude-longitude grid of finer resolution (see Horizontal Resolution). Exchanges from the surface to the atmosphere are calculated by area-averaging within the coarser atmospheric Gaussian grid, while bilinear interpolation is used for atmosphere-to-surface exchanges. Atmospheric advection of water vapor (and, on option, other tracers) is via semi-Lagrangian transport (SLT) on the Gaussian grid using cubic interpolation in all directions with operator-splitting between horizontal and vertical advection (cf. Williamson and Rasch 1989 and Rasch and Williamson 1990).

Horizontal Resolution

Spectral rhomboidal 15 (R15), roughly equivalent to 4.5 x 7.5 degrees latitude-longitude. The spectral orography (see Orography) is present at the same resolution, but other surface characteristics and variables are prescribed or calculated on a uniform 2 x 2-degree latitude-longitude grid. See also Horizontal Representation.

dim_longitude*dim_latitude:48*40 for AGCM

dim_longitude*dim_latitude: 180*90 for surface

Vertical Domain

Surface to ??? hPa; for a surface pressure of 1000 hPa, the lowest atmospheric level is at 991 hPa and the uppermost level is at 9 hPa.

Vertical Representation

Finite-difference sigma coordinates are used for all atmospheric are employed. Energy-conserving vertical finite-difference approximations are utilized, following Williamson (1983 , 1988).See also Horizontal Representation and Diffusion.

Vertical Resolution

There are 12 unevenly spaced sigma-coordinate in the vertical with the following levels: 0.009, 0.025, 0.060, 0.110, 0.165, 0.245, 0.355, 0.500, 0.664, 0.811, 0.926, 0.991 (or, for water vapor, hybrid sigma-pressure levels--see Vertical Representation). For a surface pressure of 1000 hPa, 3 levels are below 800 hPa and 5 levels are above 200 hPa.

Computer/Operating System

The PMIP simulations were run on Cray Y/MP ??? computers in a UNICOS??? environment.

Computational Performance

For the PMIP experiment, the model used about 2 minutes of Cray Y/MP computer time per simulated day.


The PMIP experiments were started from a previous multi-decadal present-day simulation and were spun up for at least 2 decades.

Time Integration Scheme(s)

Time integration is by a semi-implicit Hoskins and Simmons (1975) scheme with an Asselin (1972) frequency filter. The time step is 30 minutes for dynamics and physics, except for full radiation calculations. The longwave fluxes are calculated every 30 minutes, but with absorptivities/emissivities updated only once every 24 hours. Shortwave fluxes are computed at 1.5-hour intervals. See also Radiation.


Orography is area-averaged (see Orography). Because of the use of the SLT scheme for transport of atmospheric moisture (see Horizontal Representation), spurious negative specific humidity values do not arise, and moisture filling procedures are therefore unnecessary.

Sampling Frequency

For the PMIP simulation, monthly averages of model variables are saved.

Dynamical/Physical Properties

Atmospheric Dynamics

Primitive-equation dynamics are expressed in terms of vorticity, divergence, potential temperature, specific humidity, and the logarithm of surface pressure.


In the model troposphere, there is linear biharmonic (Ñ4) horizontal diffusion of vorticity, divergence, temperature, and specific humidity. In the model stratosphere (top three vertical levels), linear second-order (Ñ2) diffusion operates, and the diffusivities increase with height. The horizontal dvection of all fields is on constant sigma surfaces (see Vertical Representation).

The vertical diffusion of heat, momentum, and moisture is simulated by the explicit modeling of subgrid-scale vertical plumes (see Planetary Boundary Layer and Surface Fluxes).

Gravity-wave Drag


Solar Constant/Cycles

For the PMIP experiments, the solar constant is the AMIP-prescribed value of 1365 W/(m2). The orbital parameters and seasonal insolation distribution are calculated after PMIP recommendations. Both seasonal and diurnal cycles in solar forcing are simulated.


The carbon dioxide concentration is the PMIP-prescribed value of 276 ppm for control and 197 ppm for 21 cal. Zonally symmetric ozone concentrations are prescribed versus latitude, pressure level and season (as for CCM1; Bath et al, 1991).


Shortwave radiation is treated by a modified Thompson et al. (1987) scheme in ultraviolet/visible (0.0 to 0.90 micron) and near-infrared (0.90 to 4.0 microns) spectral bands. Gaseous absorption is calculated from broadband formulas of Ramanathan et al. (1983) , with ultraviolet/visible absorption by ozone and near-infrared absorption by water vapor, oxygen, and carbon dioxide treated. Reflectivities from multiple Rayleigh scattering are determined from a polynomial fit in terms of the gaseous optical depth and the solar zenith angle. A delta-Eddington approximation is used to calculate shortwave albedos and transmissivities of aerosol (see Chemistry) and of cloudy portions of each layer. Cloud optical properties depend on liquid water content (LWC), which is predicted as a prognostic variable (see Cloud Formation). Clouds that form in individual layers (see Cloud Formation) are assumed to be randomly overlapped in the vertical. The effective cloud fraction depends on solar zenith angle (cf. Henderson-Sellers and McGuffie 1990) to allow for the three-dimensional blocking effect of clouds at low sun angles.

Longwave radiation is calculated in 5 spectral intervals (with wavenumber boundaries at 0.0, 5.0 x 104, 8.0 x 104, 1.0 x 105, 1.2 x 105, and 2.2 x 105 m-1). Broadband absorption and emission by water vapor (cf. Ramanathan and Downey 1986), carbon dioxide (cf. Kiehl and Briegleb 1991), and ozone (cf. Ramanathan and Dickinson 1979) are included. In addition, there is explicit treatment of individual greenhouse trace gases (methane, nitrous oxide, and chlorofluorocarbon compounds CFC-11 and CFC-12: cf. Wang et al. 1991a , b). Cloud emissivity depends on prescribed LWC (see above). See also Cloud Formation.


Dry and moist convection as well as vertical mixing in the planetary boundary layer (PBL) are treated by an explicit model of subgrid-scale vertical plumes following the approach of Kreitzberg and Perkey (1976) and Anthes (1977) , but with simplifications. A plume may originate from any layer, and accelerate upward if buoyantly unstable; the plume radius and fractional coverage of a grid box are prescribed as a function of height. Mixing with the large-scale environmental air (entrainment and detrainment) is proportional to the plume vertical velocity. From solution of the subgrid-scale plume model for each vertical column, the implied grid-scale vertical fluxes, latent heating, and precipitation are deduced. Convective precipitation forms if the plume air is supersaturated. See also Planetary Boundary Layer.

Cloud Formation

Cloud formation follows a modified Slingo and Slingo (1991) scheme that accounts for convective, anvil cirrus, and stratiform cloud types. In a vertical column, the depth of convective cloud is determined by the vertical extent of buoyant plumes (see Convection), and the cloud fraction from a function of the instantaneous convective precipitation rate. (The convective cloud fraction is adjusted in accord with the assumption of random vertical overlap of cloud--see Radiation). If the convective cloud penetrates higher than a sigma level of about 0.6, anvil cirrus also forms.

The fraction of stratiform (layer) cloud is a function of the relative humidity excess above a threshold that depends on sigma level. In order to predict realistic amounts of stratus cloud in winter polar regions, a further constraint on cloud formation in conditions of low absolute humidity is added, following Curry and Herman (1985).


Precipitation forms in association with subgrid-scale supersaturated convective plumes (see Convection). Under stable conditions, precipitation also forms to restore the large-scale supersaturated humidity to its saturated value. The amount of evaporation is parameterized as a function of the large-scale humidity and the thickness of the intervening atmospheric layers.

Planetary Boundary Layer

Vertical mixing in the PBL (and above the PBL for an unstable vertical lapse rate) is simulated by an explicit model of subgrid-scale plumes (see Convection) that are initiated at the center of the lowest model layer using scaled perturbation quantities from the constant-flux region immediately below (see Surface Fluxes). The plume vertical motion and perturbation temperature, specific humidity, and horizontal velocity components are solved as a function of height. The implied grid-scale fluxes are then used to modify the corresponding mean quantities.


Raw orography obtained from the 1 x 1-degree topographic height data of Gates and Nelson (1975) is area-averaged over each atmospheric grid box (see Horizontal Resolution).


The ocean is represented by a thermodynamic slab (Thompson and Pollard, 1997), which crudely captures the seasonal heat capacity of the ocean mixed layer. The thickness of the slab is 50 m. Poleward oceanic heat transport is prescribed as a zonally symmetric function of latitude based on present-day observations, using 0.3 times the "0.5 x OCNFLX" case of Covey and Thompson (1989), which improves the simulation of present zonal-mean sea-surface temperatures (SSTs). Convergence under sea ice is weighted towards 2 W m-2 for 100% cover in the Northern Hemisphere, and towards 6 W m-2 in the Southern Hemisphere. To avoid unrealistic sea-ice formation in the Norwegian Sea region we impose a crude local flux that warms the mixed layer whenever it drops below 1.04 deg C, in a rectangular region between 66 and 78 deg N and -10 and 56 deg E. This flux increases linearly to a maximum possible value of 500 W m-2 if the ocean were to cool to its freezing point (-1.96 deg C). This is meant to simulate the buffering effect of the deepening winter mixed layer, and advection by warm ocean currents, and does produce wintertime heat convergences of about 200 W m-2 in agreement with Hibler and Bryan (1987). After making the sea-ice and Norwegian Sea adjustments at each time step, an additive global adjustment is made to ensure that the global integral of the convergence is zero.

Sea Ice

A three-layer sea-ice thermodynamic model predicts the local melting and freezing of sea ice, essentially as in Semtner (1976). Fractional areal cover is included as in Hibler (1979) and Harvey (1988). Sea ice advection is included using the ``cavitating-fluid'' model of Flato and Hibler (1990, 1992) in which the ice resists compressive stresses but offers no resistance to divergence or shear. The surface wind and ocean current fields for driving the dynamic sea ice were prescribed (from climatology for the winds and from an earlier OGCM run for the currents).

Snow Cover

Precipitation falls as snow if the surface air temperature is < 0 degrees C, with accumulation on land and continental/sea ice surfaces. Snow cover is simulated by a three-layer model (top layer a constant 0.03 m thick, other layers of equal thickness at each time step). Prognostic variables include the layer temperatures and the total snow mass per unit horizontal area (expressed as snow thickness and fractional coverage in a model grid box). When snow falls in a previously snow-free grid box, the fractional coverage increases from zero, with total snow thickness fixed at 0.15 m. If snowfall continues, the fractional coverage increases up to 100 percent, after which the snow thickness increases (the reverse sequence applies for melting of a thick snow cover).

Heat diffuses linearly with temperature within and below the snow. The upper boundary condition is the net balance of surface energy fluxes, and the lower condition is the net heat flux at the snow-surface interface (see below). If the temperature of any snow layer becomes 0 degrees C, it is reset to 0 degrees C, snow is melted to conserve heat, and the meltwater contributes to soil moisture. Snow cover is also depleted by sublimation (a part of surface evaporation--see Surface Fluxes), and snow modifies the roughness and the albedo of the surface (see Surface Characteristics).

The fractional coverage of snow is the same for both bare ground and lower-layer vegetation (see Land Surface Processes). In order to exactly conserve heat, temperatures are kept separately for buried and unburied lower-layer vegetation, and are adjusted calorimetrically as the snow cover grows/recedes. Any liquid water or snow already intercepted by the vegetation canopy that becomes buried is immediately incorporated into the lowest snow layer. The buried lower vegetation is included in the vertical heat diffusion equation as an additional layer between the soil and the snow, with thermal conduction depending on the local vegetation fractional coverage and leaf /stem area indices. See also Sea Ice.

Surface Characteristics

The land surface is subdivided according to upper- and lower-story vegetation (trees and grass/shrubs) of 12 types. Vegetation attributes (e.g., fractional cover and heights, leaf and stem area indices, leaf orientation, root distribution, leaf/stem optical properties, and stomatal resistances) are specified from a detailed equilibrium vegetation model driven by present-day climate. Soil hydraulic properties and albedos are prescribed as in BATS from global maps of soil-texture class and color (Dickinson et al., 1986). See also Land Surface Processes.

The surface roughness length is a uniform 2.4 x 10-3 m over the oceans and 1.0 x 10-2 m over bare soil, ice and snow surfaces.

The ocean surface albedo is specified after Briegleb et al. (1986) to be 0.0244 for the direct-beam component of radiation (with sun overhead), and a constant 0.06 for the diffuse-beam component; the direct-beam albedo varies with solar zenith angle, but not spectral interval. The albedo of ice surfaces depends on the topmost layer temperature (to account for the lower albedo of melt ponds). For temperatures that are < -5 degrees C, the ice albedos for the ultraviolet/visible and near-infrared spectral bands (see Radiation) are 0.7 and 0.4 respectively. There is no dependence on solar zenith angle or direct-beam vs diffuse-beam radiation. Following Maykut and Untersteiner (1971) , a fraction 0.17 of the absorbed solar flux penetrates and warms the ice to an e-folding depth of 0.66 m (see Sea Ice). The background albedos of land and ice surfaces are also modified by snow (see Snow Cover). The snow albedo depends on the temperature (wetness) of the topmost snow layer: below -15 degrees C, the visible and near-infrared albedos are 0.9 and 0.6, respectively; these decrease linearly to 0.8 and 0.5 as the temperature increases to 0 degrees C (cf. Harvey 1988). The direct-beam snow albedo also depends on solar zenith angle (cf. Briegleb and Ramanathan 1982).

Longwave emissivities of all solid body surfaces (including, ocean, ice, and land) are unity (blackbody emission).

Surface Fluxes

Surface solar absorption is determined from albedos, and longwave emission from the Planck equation with prescribed surface emissivities (see Surface Characteristics).

Turbulent vertical eddy fluxes of momentum, heat, and moisture are expressed as bulk formulae, following Monin-Obukhov similarity theory. The values of wind, temperature, and humidity required for the bulk formulae are taken to be those at the lowest atmospheric level (sigma = 0.993), which is assumed to be within a constant-flux surface layer. The bulk drag/transfer coefficients are functions of roughness length (see Surface Characteristics) and stability (bulk Richardson number), following the method of Louis et al. (1981). Over vegetation, the turbulent fluxes are mediated by a Land-Surface-Transfer (LSX) model (see Land Surface Processes). The bulk formula for the surface moisture flux also depends on the surface specific humidity, which is taken as the saturated value over ocean, snow, and ice surfaces, but which otherwise is a function of soil moisture.

Above the surface layer, the turbulent diffusion of momentum, heat, and moisture is simulated by a subgrid-scale plume model (see Planetary Boundary Layer).

Land Surface Processes

Effects of interactive vegetation are simulated by the LSX model (cf. Pollard and Thompson 1994 , Thompson and Pollard 1995), which includes canopies in upper (trees) and lower (grasses/shrubs) layers. Prognostic variables are the temperatures of upper-layer leaves and stems and of combined lower-layer leaves/stems, as well as the stochastically varying rain and snow intercepted by these three components (see Precipitation and Snow Cover). The LSX model also includes evaporation of canopy-intercepted moisture and evapotranspiration via root uptake, as well as soil wilting points. Air temperatures/specific humidities within the canopies are determined from the atmospheric model and the surface conditions; canopy aerodynamics are modeled using logarithmic wind profiles above/between the vegetation layers, and a simple diffusive model of air motion within each layer. Effects of vegetation patchiness on radiation and precipitation interception are also included.

Soil temperature and fractional liquid water content are predicted in 6 layers with thicknesses 0.05, 0.10, 0.20, 0.40, 1.0, and 2.5 meters, proceeding downward. (The near-surface temperature profile of the continental ice sheets is predicted by the same model.) Heat diffuses linearly, but diffusion/drainage of liquid water is a nonlinear function of soil moisture (cf. Clapp and Hornberger 1978). Boundary conditions at the bottom soil level include zero diffusion of heat and liquid, but nonzero gravitational drainage (deep runoff). The upper boundary condition for heat is the net energy flux at the soil surface computed by the LSX model; infiltration of moisture is limited by the downward soil diffusion to the center of the upper layer, assuming a saturated surface (cf. Abramopoulos et al. 1988).

Soil fractional ice content is also predicted. (Ice formation affects soil hydraulics by impeding water flow, and soil thermodynamics by changing the heat capacity/conductivity and by releasing latent heat.) The specific humidity at the upper surface of the top soil layer (used to predict evaporation--see Surface Fluxes) varies as the square of the composite liquid/ice fractions. See also Snow Cover and Surface Characteristics.

Last update November 9, 1998. For further information, contact: Céline Bonfils ( )