PARENT SESSION
Poster Session 17: Modelling
Wednesday, August 10, 5:00 PM - 6:30 PM, Exhibit Hall 220 A-E, Level 2, Palais des congrès de Montréal

Modeling CO2 Flux with NDVI and PRI in a Dynamic Chaparral.

Claudio, Helen^*,1, Cheng, Yufu¹, Gamon, John¹, Sims, Daniel², Oechel, Walter³, Luo, Hongyan³, ¹ California State University, Los Angeles, Los Angeles, CA, USA² Ball State University, Muncie, IN, USA³ San Diego State University, San Diego, CA, USA

ABSTRACT- A statistical modeling approach using optical remotely sensed data coupled with CO₂ flux data was performed at Sky Oaks Biological Field station, a chaparral-dominated ecosystem in southern California. This study focused on the data collected between October 2001 and July 2003, where the spectral data were collected with a semi-automated 100-meter tram system installed at the site and the flux data were collected with an adjacent eddy flux tower. The study period covered a normal (2001), an extreme drought (2002), and a recovery (2003) year. Simple and multiple regressions were taken with the CO₂ flux as the dependent variable to determine the extent of the effect each of the following variables: normalized difference vegetation index (NDVI), water band index (WBI), and photochemical reflectance index (PRI). For the overall study period, the strongest correlation of the three optical variables with CO₂ flux was with NDVI and PRI, indicating the importance of both canopy structure and pigment concentration in determining CO₂ fluxes. The strength and extent of the relationship between the indices and CO₂ flux were weaker following severe drought and rebounded with recovery, indicating a complicating effect of canopy structure on this physiological model. These results provide the first steps in constructing a more robust light-use efficiency model by not only identifying key components controlling CO₂ flux but also identifying how ecosystem succession and condition affect these component model terms over time.

Key words: WBI, carbon flux, chaparral, light-use efficiency model