|
Tropical forest phenology: integrated analysis of tower-mounted camera images and tower-derived GPP flux confirms satellite-based remote sensing of dry season “green up” in Amazonian forests
Jin
Wu, Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ, USA
Natalia
Restrepo-Coupe, Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ, USA
Piyachat
Ratana, Plant Functional Biology & Climate Change Cluster, University of Technology Sydney, NSW, Australia
Matt
Hayek, School of Engineering and Applied Sciences, Harvard University, Boston, MA, USA
Suelen
Marostica, National Institute for Amazon Research (INPA), Manaus, Brazil
Kenia
Wiedmann, Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ, USA
Rodrigo
da Silva, Department of Environmental Physics, University of Western Para-UFOPA
Bruce
Nelson, National Institute for Amazon Research (INPA), Manaus, Brazil
Dennis
Dye, U.S. Geological Survey
Alfredo
Huete, Plant Functional Biology & Climate Change Cluster, University of Technology Sydney, NSW, Australia
Scott
Saleska, Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ, USA, saleska@email.arizona.edu
(Presenter)
Seasonality of photosynthetic metabolism is a central topic of global change ecology, recently much-investigated by satellite-based remote sensing. In tropical Amazonian forests, these studies showed wide-scale increases in indices of photosynthesis (e.g. MODIS EVI) during the dry season, a contrast to vegetation models which represented evergreen tropical forests as water-limited, with dry-season declines in photosynthesis. However, such observations are controversial because satellite-observed patterns might arise from contamination by clouds or from seasonal changes in sun angles rather than actual changes in vegetation. We tested this “contamination hypothesis”, first, by linking surface tower-mounted remote sensing imagery (which removes effects of the atmosphere) to eddy flux tower derived GPP fluxes (in the Tapajos National Forest of Brazil), and second, by modeling the effects of seasonal changes in solar zenith angle (SZA) on the MODIS signal. We also used the tower mounted camera to directly test effects of different SZA (at different times within days) when vegetation was otherwise fixed.
We analyzed two years (2010-2011) of images from a tower mounted 3-channel (red, green, and near-infrared) TetraCAM ADC camera. A new approach based on classical image classification divided images into three components: leaves, bare wood, and open space. Camera-based phenology was combined with eddy fluxes to quantify the effect of canopy-scale phenology on ecosystem metabolism.
The satellite-based modeling study found a small effect of SZA on the MODIS-detected EVI seasonality, insufficient to change the overall pattern of dry-season green-up. This finding was confirmed by the tower-based camera study of within-day changes in SZA. The camera analysis also detected a strong dry-season “green up” (increase in camera derived leaf fraction), highly correlated (R2=0.81, p=0.001) with eddy covariance derived photosynthetic light-use efficiency. Together, these results suggest that dry-season increases in photosynthetic performance are detectable with simple camera systems, are consistent with both ground-based direct measures and with previously observed patterns from satellites, and hence, that seasonal variability in ecosystem photosynthetic metabolism is important even in evergreen tropical systems.
Presentation Type: Poster
Session: Poster Session 2-A
(Wed 11:00 AM)
Associated Project(s):
Poster Location ID: 89
|