Can satellite images track the seasonal dynamics of leaf emergence, leaf aging and litterfall in seasonal tropical evergreen forests?
Xiangming
Xiao, University of New Hampshire, xiangming.xiao@unh.edu
(Presenting)
Qingyuan
Zhang, University of New Hampshire, qzhang@eos.sr.unh.edu
Lucy
Hutyra, Harvard University, lhutyra@fas.harvard.edu
Scott
R.
Saleska, Harvard University, saleska@fas.harvard.edu
Plínio
Barbosa de
Camargo, Universidade de Sao Paulo, pcamargo@cena.usp.br
Stephen
Boles, University of New Hampshire, stephen.boles@unh.edu
Steven
C.
Wofsy, Harvard University, scw@io.harvard.edu
Berrien
Moore III, University of New Hampshire, b.moore@unh.edu
Seasonal dynamics of net ecosystem exchange of CO2 between tropical forests and the atmosphere is determined by the seasonal dynamics of photosynthesis and ecosystem respiration. A recent CO2 eddy flux study suggested that seasonal wet-dry tropical evergreen forest had high gross primary production (GPP) in late dry season (Saleska et al., 2003). Our hypothesis is that the forest evolves two adaptive mechanisms (deep roots and leaf phenology) under the environment of large seasonal dynamics of light (Xiao et al., 2004). Tropical evergreen forests have abundant species composition and complex structure, which makes it difficult to conduct ground observations of leaf phenology. Advanced optical sensors have provided systematic and consistent observations on the Earth since 1998. Here we investigated the potential of satellite images for tracking the seasonal dynamics of leaf phenology (new leaf emergence, leaf aging and leaf-fall) of seasonal tropical evergreen forest. We used 10-day composite images from the VEGETATION sensor onboard SPOT-4 satellite, and 8-day composite images from MODIS sensor onboard Terra satellite. We calculated Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI) and Land Surface Water Index (LSWI). We compare NDVI, EVI, and LSWI for the forest flux tower sites in the Amazon. Litterfall data are used to correlate with vegetation indices. We will also report a basin-wide analysis of vegetation indices with an objective to provide improved geospatial dataset of leaf phenology in Amazon basin, which would facilitate scaling-up of CO2 flux data from eddy flux tower sites and evaluating process-based biogeochemical models.
Submetido por Xiangming Xiao em 18-MAR-2004
Tema Científico do LBA: CD (Armazenamento e Trocas de Carbono)