MEASUREMENTS OF CARBON-DIOXIDE ON A VERY TALL TOWER

We present a continuous, 2-year long record of carbon dioxide (CO2) mi xing ratio at three altitudes up to 496 m above the ground on a televi sion transmitter tower in the southeastern United States. The data sho w strong diurnal and seasonal variations, and large vertical gradients . The diurnal cycles are modulated by surface uptake and release by ve getation and soils, emissions from fossil fuel combustion, and by the diurnal development of the planetary boundary layer. Gradients of 1-2 ppm between 496 m and 51 m are typically observed during summertime af ternoons, due to vigorous photosynthetic uptake. With increasing altit ude the magnitude of the diurnal cycle is damped, and daily average mi xing ratios decrease, caused by coincident changes in the sign and mag nitude of the surface flux, and changes in vertical stability of the b oundary layer over the course of the day. Measurements at 496 m give a n approximate measure (within a few tenths of a ppm) of the afternoon mean mixing ratio in the convective boundary layer. Vertical gradients between 51 m and 496 m are typically close to zero, and monthly mean mixing ratios increase slowly between November and April, indicating t hat biological activity is minimal during this period. The amplitude o f the seasonal cycle of CO2 is larger at the tower site than at marine boundary layer and mountaintop sites which are at nearly the same lat itude, because of the proximity of the tower site to terrestrial sourc es and sinks. Comparison of our continental tower data With data from ''background'' sites should provide a strong constraint for regional a nd global models of terrestrial CO2 fluxes. We also present data from weekly flask samples taken from the 496 m level and which have been an alyzed for methane (CH4), carbon monoxide (CO), and the stable isotope s of carbon in CO2 (delta(13)C). The flask data provide further inform ation about the processes that drive observed changes in CO2 mixing ra tio at the tower.