A three-dimensional model of atmospheric CO2transport based on observed winds: 1. Analysis of observational data

Abstract

Temporal and spatial patterns of atmospheric carbon dioxide elucidate the global carbon cycle as it functions on time scales ranging from days to decades. In preparation for interpreting these patterns with a three-dimensional model of atmospheric tracer transport, we have summarized CO2 measurements obtained by the Scripps Institution of Oceanography since 1957 from an array of stations between the Arctic Basin and the South Pole. Ten stations contributed to the array, supplemented by sampling on ships and ice floes. After applying consistent calibrating criteria to the full set of data, we have decomposed each record into an annually periodic signal and a seasonally adjusted time series, smoothed to emphasize interannual variations. We have computed harmonic coefficients to express the average seasonal cycle. We have determined seasonally adjusted concentrations for 1962, 1968, and annually from 1978 to 1986, to reflect slowly varying characteristics of the carbon cycle, and we have assembled these data in north-to-south profiles to reveal spatial patterns. We have similarly assembled isotopic data derived from measurements of the 13C/12C isotopic ratio of atmospheric CO2, made always on the same air measured for its CO2 concentration. We have extended the data sets for stations at Mauna Loa Observatory, Hawaii and the South Pole through 1988 to create a continuous time series which approximates the global change in CO2 concentration over 32 years and the isotopic ratio over 11 years. To establish interannual changes in the carbon cycle, we have developed a compartment model which treats the transfers of carbon between global atmospheric and terrestrial biospheric carbon pools, and between these pools and a world ocean in which vertical transport occurs by diffusion. In a variant to this model the oceanic submodel was replaced with a three-dimensional oceanic transport model. Both the concentration and isotopic ratio of CO2 show clearly defined seasonal cycles and evidence that the carbon cycle responds to El Niño events that recur in the records approximately every 4 years. According to the isotopic data, oscillations in CO2 associated with El Niño events are produced by opposing oceanic and biospheric fluxes several times larger than the net fluxes inferred from the CO2 concentration data alone. On a longer time scale the concentration data show a weak, approximately 11 year, cycle possibly driven by variations in solar irradiance. On a still longer time scale the data indicate that a larger apparent fraction of CO2 from fossil fuel combustion has been retained in the air during the past 14 years than formerly, in spite of a reduced acceleration in worldwide fuel consumption which, according to the compartment model, should have led to a lesser retention in the air after 1974. The isotopic records suggest that this increased retention is partially a result of ocean warming, but is predominantly caused by an accelerated release of CO2 by the terrestrial biosphere.