Measuring the spcetral expression of carbon dioxide in the solar reflected spectrum with AVIRIS

Carbon dioxide is a low-concentration, but important, component of the Earth's atmosphere. This gas absorbs electromagnetic radiation (EMR) in several regions of the spectrum. Absorption of energy by carbon dioxide adds heat to the atmosphere. In the world today, the burning of fossil fuels and other anthropogenic processes adds carbon dioxide to the atmosphere. Other natural processes in the Earth's system both add and remove carbon dioxide. Overall, measurements of atmospheric carbon dioxide at selected sites around the globe show an increased carbon dioxide concentration in the atmosphere. A figure shows the measured carbon dioxide from Mauna Loa, Hawaii, from 1958 to 2000. Overall, the concentration has increased from 315 to 365 ppm at this site over this period. (There is also a yearly cycle to the concentration that is timed with and hypothesized to be related to the vegetation growing season in the Northern Hemisphere.) The overall expected effect of this increase of atmospheric carbon dioxide is trapping of heat in the atmosphere and global warming. While this overall relationship between carbon dioxide and global warming seems straightforward, many of the specific details relating to regional and local sources and sinks and gradients of carbon dioxide are not well understood. A remote sensing capability to measure carbon dioxide could provide important inputs for scientific research to better understand the distribution and change in atmospheric carbon dioxide at detailed spatial and temporal levels. In pursuit of this remote sensing of carbon dioxide objective, this paper analyzes the expression of carbon dioxide in the spectral range measured by the Airborne Visible/Infrared Imagery Spectrometer (AVIRIS). Based on these analyses, a spectral-fitting algorithm that uses AVIRIS measured spectra and MODTRAN radiative-transfer code modeled spectra to derive total column carbon dioxide abundance has been developed. This algorithm has been applied to an AVIRIS data set acquired over Pasadena, California, in 1999 and a data set acquired over the Pacific Ocean near Hawaii in 2000 with promising results. This is ongoing research; the current initial analyses, measurements, and results are reported in this paper.