Remote Sensing of Ecosystem Light Use Efficiency Using MODIS
Karl
Fred
Huemmrich, UMBC, karl.f.huemmrich@nasa.gov
(Presenting)
Elizabeth
M
Middleton, NASA/GSFC, elizabeth.m.middleton@nasa.gov
David
R
Landis, Sigma Space, david.r.landis@gsfc.nasa.gov
T
Andrew
Black, UBC, andrew.black@ubc.ca
Alan
G
Barr, Climate Research Branch, Meteorological Service of Canada, alan.barr@ec.gc.ca
J
Harry
McCaughey, Queens University, mccaughe@post.queensu.ca
Forrest
G
Hall, UMBC, fghall@ltpmail.gsfc.nasa.gov
Understanding the dynamics of global carbon cycle requires a determination of spatial and temporal distributions of photosynthetic CO2 uptake by vegetation. Optimal photosynthetic function is affected by stress factors causing down-regulation (i.e., reduced photosynthetic rates). Present modeling approaches of ecosystem carbon exchange rely on meteorological data inputs to predict the relative photosynthetic function in response to environmental conditions inducing stress (e.g., drought, high/low temperatures). This study examines the determination of ecosystem light use efficiency (LUE) from remote sensing, through measurement of vegetation spectral reflectance changes associated with physiologic stress responses exhibited by photosynthetic pigments. We use Moderate-Resolution Spectroradiometer (MODIS) on Aqua and Terra to provide frequent, narrow-band measurements. The MODIS ocean bands were used to calculate the Photochemical Reflectance Index (PRI), an index sensitive to reflectance changes near 531nm associated with vegetation stress responses by pigments in the xanthophyll cycle. MODIS PRI values were compared with LUE from CO2 fluxes measured at four Canadian forest sites: mature Douglas fir site in British Columbia, mature aspen and black spruce sites in Saskatchewan, and mixed forest site in Ontario, all part of the Canadian Carbon Program network. The relationships between LUE and MODIS PRI were different among forest types, with clear differences in the slopes of the relationships for conifer and deciduous forests. The MODIS based LUE measurements provide a more accurate estimation of observed LUE than the values calculated using the existing MODIS GPP model. This suggests the possibility of a GPP model that uses MODIS LUE instead of modeled LUE. This type of model may provide a useful contrast to existing models driven by meteorological data. The main impediment to developing such a model is the lack of a MODIS product providing surface reflectance for MODIS ocean bands over land.
Presentation Type: Poster
Poster Session: Carbon Cycle Science
NASA TE Funded Awards Represented:
Middleton, Elizabeth
Spectral Bio-Indicators of Ecosystem Photosynthetic Efficiency