CR 28:11-22 (2004)  -  doi:10.3354/cr028011

Correlation between atmospheric CO2 concentration and vegetation greenness in North America: CO2 fertilization effect

C. Lim1,*, M. Kafatos1, P. Megonigal2

1Center for Earth Observing and Space Research (CEOSR), School of Computational Sciences (SCS), George Mason University, Fairfax, Virginia 22030, USA
2Smithsonian Environmental Research Center, PO Box 28, 640 Contees Wharf Road, Edgewater, Maryland 21037, USA
*Email:

ABSTRACT: The possibility that rising atmospheric CO2 concentrations are influencing plant growth in contemporary ecosystems has received little attention, and the studies that exist have been done on a small spatial scale. We correlated the monthly rate of relative change in normalized differenced vegetation index (NDVI) from advanced very high resolution radiometer (AVHRR) data to the rate of change in atmospheric CO2 concentration during the natural vegetation growing season for evidence of a possible CO2 fertilization effect on vegetation development. The study addressed seasonal and annual patterns in spatially averaged NDVI for 3 different ecological regions in North America from 1982 to 1992. Correlations between CO2 and NDVI were calculated for 3 different lag conditions. Relatively high and positive correlation coefficients were found when the monthly rate of change in NDVI was 1 mo lagged to that for CO2, which suggests, but does not prove, a CO2 fertilization effect on natural vegetation development. Generally, the correlation coefficients changed from relatively high and positive correlations when NDVI was lagged 1 mo behind CO2 to relatively high and negative correlations when CO2 was lagged 1 mo behind NDVI. A general increase in the annual maximum greenness of the vegetation was also found in most of the regions studied from 1982 to 2001. The desert and humid temperate regions in the eastern part of North America showed an increase in the annual minimum vegetation greenness, while the southern humid temperate regions showed relatively high correlations between the minimum NDVI and atmospheric CO2 concentration in interannual comparisons. The results of this study are generally consistent with the notion of a contemporary CO2 fertilization effect, but they also demonstrate how remotely-sensed data can be used to explore the effects of global change at large scales in order to complement experimental results obtained on smaller temporal and spatial scales.


KEY WORDS: Global change · Climate change · CO2 fertilization effect · NDVI · Remote sensing · Vegetation


Full article in pdf format