GRADUATE STUDIES BULLETIN
ANNOUNCEMENT

Florida International University
Division of Graduate Studiesivision of Graduate Studies

Thesis Proposal Seminar

Abstract

Relating seasonal patterns of CO₂ flux to spatial representations of Normalized Difference Vegetation Index (NDVI) in Arctic Alaska.

by

Inga Parker

Climate change in the Arctic will differentially affect leaf phenology and species composition of tundra, resulting in changing patterns and magnitudes of ecosystem CO₂ flux. Seasonal patterns of CO₂ flux are a function of the combined patterns of leaf phenology and species contributions to the ecosystem, but ascribing flux patterns to individual species or functional groups is problematic. As part of the International Tundra Experiment (ITEX) program, we have been measuring CO₂ flux and the Normalized Difference Vegetation Index (NDVI) on growing season manipulations to investigate Gross Primary Productivity (GPP) and relative NDVI values independently. Ecosystem CO₂ fluxes from within permanent chamber bases and NDVI images of the chamber base areas from an Agricultural Digital Camera (ADC) provide an opportunity to directly compare these ecosystem parameters. Digital images from the ADC allow evaluation of NDVI for specific areas of the image as opposed to the traditional field spectroradiometer, which provides only a single value or an average of several small footprints. Preliminary oblique NDVI images from our study site have been analyzed with promising results, showing a significant correlation between NDVI and CO₂ flux. I am proposing to analyze images photographed from nadir for area-specific correlations between NDVI and CO₂ flux and compare them with areal point frame data coverage which includes individual species and functional groups. My primary hypothesis is that changes in specific functional groups are responsible for the major seasonal changes in NDVI/CO₂ flux within and among season manipulation treatments. The final product should present usable regression equations and rapid Geographic Information System (GIS) evaluations relating CO₂ flux to NDVI as well as the proportion of functional groups contributing to the greatest changes in seasonal fluxes. The widely used spectral bands of NDVI also create the potential for introducing usable model parameters for scaling from plots to larger spatial areas.

Seasonal changes in ecosystem CO₂ flux result from the combined patterns of leaf phenology of individual species or functional group contributions, but ascribing flux patterns to these species or functional groups is problematic. Climate change may differentially affect leaf phenology and composition of functional groups, resulting in changing patterns of ecosystem CO₂ flux. As part of the International Tundra Experiment (ITEX) program, we have been measuring CO₂ flux and the Normalized Difference Vegetation Index (NDVI) on season manipulation treatments to investigate Gross Primary Productivity (GPP) and relative NDVI values independently. Ecosystem CO₂ fluxes measures encompassing from within permanent static chamber bases areas and NDVI images using an Agricultural Digital Camera (ADC) for the chamber base areas provide an opportunity to directly compare these ecosystem parameters. ADC digital images provide a unique abilityallow to evaluate evaluation of specific areas of the image as opposed to the traditional field spectroradiometer which provides an average of several small footprints. Preliminary oblique NDVI images from this my study site have been analyzed with promising results, showing a significant correlation between NDVI and CO₂ flux. Images photographed from nadir for specific area correlations of NDVI and CO₂ flux need to be analyzed and compared with point frame functional group data over the chamber base areas. A My main primary hypothesis is that these comparisons changes in specific functional groupswill indicate which groups are responsible for the major seasonal changes in NDVI/CO₂ flux within and among climate scenario season manipulation treatments. The final product should present be usable correlations from NDVI values derived from ADC images, usable for resulting in quick evaluations of CO₂ flux for a specified area as well as the proportion of functional groups responsible for major flux differences. The widely used spectral bands of NDVI also create the potential for introducing usable model parameters for scaling from plots to larger spatial areas.

Date: March 25, 2002 Department: Biological Sciences
Time: 11 a.m.
Place: University Park, WC130 Major Professor: Dr. Steven Oberbauer