Estimating vegetation structural effects on carbon uptake using satellite data fusion and inverse modeling

Abstract

Abstract. Regional analyses of biogeochemical processes can benefit significantly from observational information on land cover, vegetation structure (e.g., leaf area index), and biophysical properties such as fractional PAR absorption. Few remote sensing efforts have provided a suite of plant attributes needed to link vegetation structure to ecosystem function at high spatial resolution. In arid and semiarid ecosystems (e.g., savannas), high spatial heterogeneity of land cover results in significant functional interaction between dominant vegetation types, requiring new approaches to resolve their structural characteristics for regional-scale biogeochemical research. We developed and tested a satellite data fusion and radiative transfer inverse modeling approach to deliver estimates of vegetation structure in a savanna region of Texas. Spectral mixture analysis of Landsat data provided verifiable estimates of woody plant, herbaceous, bare soil, and shade fractions at 28.5 rn resolution. Using these subpixel cover fractions, a geometric-optical model was inverted to estimate overstory stand density and crown dimensions with reasonable accuracy. The Landsat cover estimates were then used to spectrally unmix the contribution of woody plant and herbaceous canopies to AVHRR multiangle reflectance data. These angular reflectances were used with radiative transfer model inversions to estimate canopy leaf area index (LAI). The suite of estimated canopy and landscape variables indicated distinct patterns in land cover and structural attributes related to land use. These variables were used to calculate diurnal PAR absorption and carbon uptake by woody and herbaceous canopies in contrasting land cover and land use types. We found that both LAI and the spatial distribution of vegetation structural types exert strong control on carbon fluxes and that intercanopy shading is an important factor controlling functional processes in spatially heterogeneous environments. IntroductionChanges in land use often lead to changes in vegetation structure, which significantly impact biosphere-atmosphere exchange processes and biogeochemical cycles (e.g., carbon fluxes, nutrient dynamics). However, structural differences (e.g., type and amount of tissue) between vegetation types also predispose different ecosystems to distinct functional responses to human disturbance and climate variability [e.g., Townsend et al. To analyze actual rather than potential ecosystem function at regional scales, especially under conditions of changing land use, observations of changing vegetation structure must be linked to functional processes. To date, our ability to connect function to structure has been limited because (1) the physiological and biogeochemical links between structure and function are highly complex and (2) observation of structure is often constrained to detailed ground studies over very small scales (e.g., meters). Studies focusing on structure at large scales (e.g., kilometers) often lack sufficient detail required for many analyses of e...

Grassland to Woodland Transitions: Integrating Changes in Landscape Structure and Biogeochemistry

“…Collectively, these studies indicate a positive relationship at regional scales between woody plant-induced increases in ecosystem C-sequestration potential and mean annual rainfall. Approaches such as the one used here, which link remote sensing, field sampling, and ecosystem modeling, are emerging as plausible means for addressing this complexity (Wessman 1992, Asner et al 1998, Akiyama and Koizumi 2002 These contrasting perspectives point to the need for a better understanding of how ecosystem primary production and SOC pools change with replacement of grasses by woody plants (e.g., House et al 2003).…”

Assessing the Spectral Properties of Sunlit and Shaded Components in Rice Canopies with Near-Ground Imaging Spectroscopy Data

“…In the community of vegetation mapping, most studies focus on the sunlit portions of the upper canopy and discard the shaded portions in subsequent spectral analysis [5,6,7,8,9,10,11]. Others assume that the reflectance values of shadows equal zero or are constant for the purposes of spectral mixture analysis [2,12] and estimation of canopy parameters with radiative transfer [13,14] or geometrical optical models [15,16,17]. These studies suggest that the potential spectral information in shaded portions of the canopy have not been fully exploited.…”

Coupling spectral unmixing and multiseasonal remote sensing for temperate dryland land-use/land-cover mapping in Minqin County, China

“…Finally, the physically based domains of EMs provide a more intuitive link between image measurements and observations in the field (Adams et al 1995), and fractions modelled by LSMA of TM images have been linked to biophysical parameters of vegetation in boreal forest and savannah ecosystems (e.g. Hall, Shimabukuro, and Huemmerich 1995;Asner et al 1998;Peddle, Brunke, and Hall 2001). Hall, Shimabukuro, and Huemmerich (1995) and Sonnentag et al (2007) found the excellent shade fraction models for mapping tree and shrub leaf area indices (LAIs).…”