Skip to main content
2023 ESA Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Back

Organized Oral Session

Hybrid Session

Session: : Linking Optical and Energy Fluxes Workshop

OOS 33-3 - From signal to mechanism in remote sensing of forests

Tuesday, August 8, 2023
4:00 PM – 4:15 PM PDT
Location: 252
Our understanding of physical and biological processes in tropical forests is limited by sparse observations of tree physiology and biosphere-atmosphere fluxes of CO2, water, and energy. Remote sensing technologies promise greater spatial coverage than flux networks in tropical forests, but linking signal to biological mechanism remains a challenge in these diverse forests. We describe an approach to quantifying spatiotemporal patterns of forest structure, physiology and fluxes (carbon and water) through integrating a biophysical model (SCOPE, the Soil Canopy Observation of Photochemistry and Energy fluxes model) with forest inventory data and current remote sensing data streams relevant to quantifying canopy properties, forest structure, and physiology. These remote sensing data include lidar from GEDI, hyperspectral reflectance from DESIS, solar-induced fluorescence from OCO-3 and TROPOMI, and land surface temperature from ECOSTRESS. SCOPE is a biophysical model designed to link remote sensing observations in the visible, infrared and thermal wavelengths with land surface processes. Comparing SCOPE simulations of fluxes (based on field data) with eddy covariance data offers a test of hypothesized processes and parameter sets for a given tropical forest site. Inverting SCOPE to estimate leaf and canopy traits (e.g. leaf pigment content) from satellite hyperspectral reflectance data enables observing how leaf, plant and ecosystem physiology changes across climatic and topographic gradients.

We chose to test our approach using pantropical sites with eddy covariance towers co-located with large inventory plots. In particular, we focused on the Smithsonian’s Forest Global Earth Observatory (ForestGEO) network of large forest plots (15 - 52 hectares) where every stem > 1 cm diameter at 1.3 m DBH (diameter-at-breast height) is identified, mapped, and measured every five years, enabling understanding of the tree species composition of remote sensing pixels, and facilitating scaling up leaf physiology data to the canopy. ForestGEO plots co-located with eddy covariance towers represent different positions on the wet/dry spectrum of moist tropical forests, including Barro Colorado Island (with significant wet/dry seasonality), and ever-wet tropical forests in Malaysia. Amazon sites (non-ForestGEO sites) with less pronounced wet/dry seasonality bridge these end members. Preliminary results show that including forest species composition is important for reasonable simulated photosynthetic fluxes using our approach. We discuss challenges to harnessing remote sensing data for process-based modeling of fluxes, as well as opportunities to identify drivers of tropical forest drought response using our approach.