The potential of the MERIS Terrestrial Chlorophyll Index for carbon flux estimation

Abstract

In this study we evaluated the potential of the Medium Resolution Imaging Spectrometer (MERIS) Terrestrial Chlorophyll Index (MTCI) for monitoring gross primary productivity (GPP) across fifteen eddy covariance towers encompassing a wide variation in North American vegetation composition. The acrosssite relationship between MTCI and tower GPP was stronger than that between either the MODIS GPP or EVI and tower GPP, suggesting that data from the MERIS can be used as a valid alternative to MODIS for estimating carbon fluxes. Correlations between tower GPP and both vegetation indices (EVI and MTCI) were similar only for deciduous vegetation, indicating that physiologically driven spectral indices, such as the MTCI, may also be able to complement existing structurally-based indices in satellite-based carbon flux modeling efforts. 1.Introduction Quantitative estimates of carbon dioxide exchange at regional to global scales are critical to improve our understanding of the links between carbon and climate. Tower-based eddy covariance (EC) techniques have been used across a wide range of ecosystems to provide information on seasonal and inter-annual carbon fluxes. However, flux tower sites only account for carbon fluxes within the designated tower footprint and the number and geographical distribution of towers across the globe is limited. Other attempts at estimating terrestrial carbon fluxes have concentrated on the development of process-based ecosystem exchange models (e.g. the Boreal Ecosystem Productivity Simulator (BEPS; Liu et al. 1997) and the Terrestrial Ecosystem Model (TEM; e.g. Raich et al. 1991)). Whilst such models show great promise, their applicability at regional and global scales is challenging due to their complexity and requirements for data that are often scarce or unavailable at the appropriate spatial and temporal scales. Carbon flux models that are driven by remotely sensed observations can be used to estimate gross primary productivity (GPP) frequently and over large areas; for example the NASA Carnegie-Ames-Stanford (NASA-CASA) model (Potter et al. 1993), the Terrestrial Uptake and Release of Carbon (TURC) model (Ruimy et al. 1996) and the Moderate Resolution Imaging Spectrometer Global Primary Productivity (MODIS-MOD17 GPP) model (Running et al. 2004)). The vast majority of satellite-based models are 'Production Efficiency Models' (PEMs) based on the light use (LUE) efficiency concept for conversion of absorbed photosynthetically active radiation (APAR) into biomass (Monteith 1977). In most PEMs the maximum LUE is empirically derived based on vegetation type and then reduced according to meteorological indicators of environmental stress. Thus whilst some PEM parameters can be estimated from satellite data, for example the fraction of absorbed photosynthetically active radiation (fPAR; Myneni et al., 2003;Prince & Goward 1995), the estimation of others, such as LUE, depend upon the availability of metrological data and vegetation maps. There can, however, be substantial errors in the...

Remote estimation of crop gross primary production with Landsat data

“…2), as well as low and high frequency variations of PAR in . Chl-related VI closely follow seasonal changes in Chl content (Almond et al, 2010;Gitelson et al, 2005;Harris & Dash, 2010;Wu et al, 2009). In the green-up stage, VI alone relate closely to GPP, while in reproductive and senescence stages, as PAR in declines, discrepancies between VI and GPP increased Sakamoto et al, 2011).…”

“…The VI-PAR in -based model (Eq. 3), with MTCI derived from MERIS images, was capable of estimating GPP accurately across a variety of land cover and vegetation types (Almond et al, 2010;Harris & Dash, 2010). Wu et al (2010aWu et al ( , 2011 showed that maize GPP could be estimated with high accuracy using the VI-PAR inbased model with MODIS data.…”

A Novel Principal Component Analysis Method for the Reconstruction of Leaf Reflectance Spectra and Retrieval of Leaf Biochemical Contents

“…The statistical-empirical method is typically based on a regression function that links measured biochemical or biophysical parameters to spectral measurements [7,8], such as spectral reflectance [9,10], spectral indices [11][12][13][14], red-edge features [15][16][17][18], absorption features [19][20][21][22], and band combinations made using the wavelet transformation or principal transformation [23][24][25][26]. However, the leaf spectrum depends on a complex interaction between internal and external factors that may vary significantly from one species to another, and it is hard to build a universal relationship between a single vegetation variable and a spectral signature [27].…”

Remote sensing-based estimation of gross primary production in a subalpine grassland

“…Based on the assumption that chlorophyll is related to the presence of photosynthetic biomass, which is essential for primary production and thus conceptually related to GPP , recent studies (Gitelson et al, 2008;Harris and Dash, 2010) suggest that GPP can be estimated through direct correlation with chlorophyll-related indexes. Successful results have been obtained in agricultural crops (Gitelson et al, 2008).…”