Model GISS (Russell): Elaborations

Participation

Spinup/Initialization

• The atmospheric model was initialized from the NMC analysis for December 1, 1977.

• The ocean model was initialized from the Levitus (1982) temperature and salinity fields.  In the upper 1500 m, the Levitus seasonal climatology was temporally interpolated to obtain values appropriate for December 1; below 1500 m, Levitus annual-mean data were used. (The Levitus climatologies on 33 vertical levels were vertically integrated to the 13 layers in the ocean model.) The initial ocean free-surface boundary condition was zero elevation everywhere (i.e., mean sea-level conditions).  All ocean currents were initially 0, but the height field adjusted soon after the start of coupled integration to conform to that for geostrophic flow.  The initial heat and salt in the model's straits were calculated by using the ocean values at the two ends of each strait.  The initial sea ice distribution was after Walsh and Johnson (1979) in the Northern Hemisphere, and after Alexander and Mobley (1976) in the Southern Hemisphere.

• The atmospheric and ocean models were coupled and integrated for 120 years without flux adjustment, the ocean's peak heat transports becoming relatively stable after about 25 years.  Thereafter, a slow climate drift was evident in the (increasing) sea ice extents in the North Atlantic, the (weakening) thermohaline circulation, the (decreasing) elevation of the ocean's free surface, the imbalance of 3.7 W m^-2 in ocean surface radiative fluxes, and the 0.007 degrees K per year warming in the mass-weighted vertically integrated ocean temperature.

Land Surface Processes

• Soil temperature is computed by solving a heat diffusion equation in two layers, where the upper layer (depth 0.1 m) simulates diurnal fluctuations, and the lower layer (depth 4 m) seasonal variations. Thermal conductivity and heat capacity vary with snow cover and soil moisture amount and phase (i.e., liquid vs solid water). In vegetated regions during winter, the thermal conductivity also reflects the insulating effects of dead grass and leaves.
• Soil moisture, predicted in two layers, is expressed as a ratio of available water to field capacity, which is a function of surface type and vegetation according to Matthews (1983, 1984). Soil moisture in the upper layer reflects short-term variations in hydrological variables, while that of the lower layer functions as a reservoir. Surface evaporation is parameterized as the product of the fractional wetness of the upper layer and the potential evaporation. Moisture diffusion from the lower layer varies seasonally to account for greater potential for depletion of this reservoir during the growing season; otherwise, the effects of a vegetation canopy are not explicitly simulated. Runoff is represented as one-half the product of the fractional wetness of the upper layer and the rainfall rate, with additional runoff occuring as needed to keep the fractional wetness no larger than unity. Cf. Hansen et al. 1983 for further details.
• Runoff is returned to the ocean by means of a river transport model (cf. Miller et al. 1994). Runoff instantaneously increases the river and lake mass of a grid box, which is also affected by the local balance of precipitation and evaporation. For each continental grid box, a river direction file indicates the downstream routing to one of the 8 neighboring grid boxes. The mass flux out of a grid box is computed as a function of the river and lake mass that lies above the local sill depth, an effective flow speed that depends on the local orography gradient, and the mean distance to the downstream neighbor. Flow speeds are globally optimized to ensure that fresh water arrives at coastal grid boxes with the proper seasonal timing.

Sea Ice

• Sea ice forms when the uppermost ocean layer cools below the freezing point of salt water. Ice cover is determined both by energy exchange with the atmosphere and by ocean heat transport and heat capacity, following Hansen et al. (1988). Initially, the ice is 0.5 m thick; as it melts, the ice is contracted horizontally so as to maintain a 0.5 m thickness. The area of ice leads is calculated after the method of Hansen et al. (1983): the minimum fraction of open ocean area in a grid box varies inversely with the ice thickness. As sea ice forms or thickens, the salinity of the uppermost ocean layer increases.

• Snow also may accumulate on the ice, and a two-layer model determines the temperatures within the ice and the snow pack. When, from consideration of surface heat balance, the temperature begins to exceed the freshwater melt temperature of 0 deg C, it remains there as long as the ice/snow melting continues.  If the temperature of the uppermost ocean layer rises above 0 deg C, the sea ice melts both vertically and laterally, thereby keeping the ocean temperature at 0 deg C.

• Sea ice is not advected; instead, the atmospheric momentum stress on the ice is passed directly to the uppermost layer of the ocean model.

Chief Differences from Closest AMIP Model

The chief differences from this AMIP model include:

Horizontal Representation

The momentum equation is expressed on a C-grid (not on a B-grid as in the AMIP model). The linear upstream scheme of Russell and Lerner (1981) is used for the advection of heat and water vapor.

Convection

The convection scheme is after Hansen et al. (1983), not Del Genio and Yao (1988), as in the AMIP model.

Cloud Formation

A diagnostic cloud formation scheme is used instead of the Del Genio et al. (1993) prognostic formulation in the AMIP model.

Precipitation

Precipitation is determined diagnostically, rather than prognostically as in the AMIP model.

Planetary Boundary Layer

The formulation of the PBL is simpler than that of the AMIP model, in that surface air quantities are linearly related to values in the first full atmospheric layer and to their vertical gradients.

Land Surface Processes

A simple 2-layer land-surface model is used instead of the Abramopoulos et al. (1988) scheme in the AMIP model.

References

Alexander, R.C., and R.L. Mobley, 1976: Monthly average sea-surface temperatures and ice-pack limits on a 1 degrees global grid. Mon. Wea. Rev., 104, 143-148.

Hansen, J., G. russell, D. Rind, P. Stone, A. Lacis, S. Lebedeff, R. Ruedy, and L. Travis, 1983: Efficient three-dimensional global models for climatic studies: Models I and II. Mon. Weather Rev., 111, 609-662.

Hansen, J.; I. Fung, A. Lacis, S. Lebedeff, D. Rind, R. Ruedy, and G. Russell. 1988.  Global climate changes as forecast by the Giss 3-D model. J.Geophys.Res.,  92: 14739-14760.

Levitus, S., 1982: Climatological atlas of the world's oceans. NOAA Professional Paper 13, 173 pp.

Matthews, E., 1983: Global vegetation and land use: New high-resolution data bases for climate studies. J. Clim. Appl. Meteor., 22, 474-487. (Abstract)

Matthews, E., 1984: Vegetation, land-use, and seasonal albedo data sets: Documentation of archived data tape. NASA Tech. Memo. 86107, National Aeronautics and Space Administration, Washington, D.C., 20 pp.

Miller, J.R., G.L. Russell, and G. Caliri, 1994: Continental scale river flow in climate models. J. Climate, 7, 914-928.

Russell, G.L., and J.A. Lerner, 1981: A new finite differencing scheme for the tracer transport equation. J. Appl. Meteorol., 20, 1483-1498.
 

Russell, G.L., J.R. Miller, and D. Rind, 1995: A coupled atmosphere-ocean model for transient climate change studies.  Atmosphere-Ocean, 33, 683-730.

Walsh, J., and C. Johnson, 1979: An analysis of Arctic sea ice fluctuations. J. Phys. Oceanogr., 9, 580-591.
 
 

Top

CMIP Documentation Directory

Last update 15 May, 2002.  This page is maintained by Tom Phillips (phillips@pcmdi.llnl.gov).
 

LLNL Disclaimers

UCRL-MI-126431