Combining field and space borne reflectance time series for evaluation of canopy function and productivity
Petya Campbell, NASA GSFC / UMBC JCET, petya.campbell@nasa.gov (Presenter)
Karl Fred Huemmrich, NASA GSFC / UMBC, huemmric@umbc.edu
Petr Lukes, Global Change Research Institute CAS, Brno, Czech Republic, lukes.p@czechglobe.cz
Christopher Neigh, NASA GSFC, christopher.s.neigh@nasa.gov
Benjamin Poulter, NASA GSFC, benjamin.poulter@nasa.gov
Jana Albrechtova, Charles University, Prague, Czech Republic, jana.albrechtova@natur.cuni.cz
Christiaan Van Der Tol, University of Twente, tol@itc.nl
Sean M McMahon, Smithsonian Environmental Research Center, mcmahons@si.edu
High spectral resolution data provide an efficient tool for assessment of vegetation photosynthetic function. This study presents preliminary findings from analysis of field Fluorescence Box (FLoX, JB Hyperspectral) and space-borne DLR Earth Sensing Imaging Spectrometer (DESIS) reflectance time series covering the visible near-infrared regions (VNIR). Data were collected at a different temporal, spectral and spatial resolutions at the Smithsonian Environmental Research Center (SERC) deciduous forest eddy covariance site in Edgewater, MD, USA. Mid-day proximal field and space borne reflectance measurements corresponding by acquisition date were assembled and the differences in reflectance properties and canopy traits (e.g., chlorophyll, carotenoids, dry matter, and water content) were evaluated for the flux tower footprint. We used the Soil Canopy Observation Photosynthesis Energy (SCOPE) biophysical model in an inversion to estimate canopy traits and in a forward simulation to derive gross primary productivity (GPP). The DESIS seasonal and dense FLoX time series corresponded reasonably well for the same timeframe and represented the dynamics in reflectance and canopy chlorophyll associated with the seasonal variations in environmental conditions. The SCOPE estimation of canopy chlorophyll, LAI and GPP produced more accurate results with lower RMSE using FloX as compared to DESIS, which is likely due to the imperfection of the atmospheric correction to surface reflectance of the DESIS data and the stronger influence of canopy structure to the DESIS data. Additional analysis will be conducted using only the DESIS pixels from the FloX field of view to further compare the two datasets and assess the drivers of uncertainty . Comparing the proximal and space-borne estimates of vegetation traits demonstrates the feasibility of a multi-sensor approach upscaling canopy traits from field to satellite level. Our preliminary results demonstrate the utility of dense hyperspectral time series for monitoring the seasonal dynamics in vegetation function. The constellation of forthcoming spectroscopy missions, such as SBG, PACE, CHIME, etc. hold great potential to develop multi-sensor time-series that capture vegetation dynamics and enable trait comparisons across multiple seasons and years.
Poster: Poster_Campbell_1-22_222_35.pdf
Associated Project(s):
Poster Location ID: 1-22
Presentation Type: Poster
Session: Poster Session 1
Session Date: Tue (May 9) 5:00-7:00 PM
CCE Program: TE