A three-dimensional model of atmospheric CO2transport based on observed winds: 2. Model description and simulated tracer experiments

Abstract

The three-dimensional atmospheric transport model developed at the Goddard Institute of Space Sciences (GISS) has been modified in its coarse-grid version (7.83° × 10.00° horizontal resolution, 9 layers in vertical direction) by replacing the original model-generated wind fields with observed winds of the Global Weather Experiment, covering the period December 1978 through November 1979. The parameterization of subgridscale vertical convection was retained from the earlier model version, except that all intensities were reduced by 50 percent. To simulate atmospheric CO2 and its 13C/12C ratio, sources and sinks of carbon at the earth's surface were prescribed. The net primary productivity of the terrestrial biosphere was computed from vegetation index (NDVI) data representing the greenness of the land as recorded by the AVHRR instrument flown on satellites of the U.S. National Oceanic and Atmospheric Administration. The productivity for a given greenness was assumed to be proportional to the amount of photosynthetically active radiation reaching the plant canopy as computed from estimates of solar insolation under clear sky conditions with an allowance for attenuation caused by the presence of clouds. Estimates of cloud cover were based on satellite data of daily albedo. The respiration of terrestrial plant detritus and soils was assumed to be dependent on temperature but was globally adjusted to achieve an optimal fit of the model prediction of the seasonal cycle of atmospheric CO2 to that observed at four northern hemisphere stations. The oceanic exchange of CO2 was prescribed from rough estimates of the CO2 partial pressure of sea water, assuming a constant air-sea exchange rate. The mean annual partial pressure field was afterwards adjusted to achieve an optimal fit with respect to broad geographic features of the observed mean annual atmospheric CO2 field. The seasonal variation in partial pressure was assumed to depend on sea-surface temperature, but the variation was reduced poleward of 35° in both hemispheres to account, at least crudely, for marine plant activity. Perturbations of the global carbon cycle arising from industrial CO2 production and anthropogenic changes in land use were also modeled, as well as a postulated net annual uptake of CO2 by the terrestrial biosphere in recent years attributed to enhanced plant activity. The 13C/12C of atmospheric CO2 was predicted by the model taking into account the degree to which these stable isotopes were fractionated by the sources and sinks prescribed for the CO2 concentration. Purely isotopic source components arising from the temperature dependency of the air-sea exchange of CO2 and resulting from the carbon-13 Suess effect were also taken into account. In addition the transport properties of the model were tested by simulating the dispersal of two radioactive tracers having simple source/sink configurations: krypton-85 and radon-222. The model gives satisfactory predictions of the meridional concentration gradient of krypton-85 and of the seasonally changing vertical gradient of radon-222 over the continents.