Monitoring tree-grass agroforestry systems (TGA) either through in situ field observation or remote sensing, requires compilation of data
on the biochemical and structural characteristics of the (one/two) vegetation layers composing these ecosystems in order to derive reliable information about relevant management indicators such as biomass, quality, phenological stage, productivity, species composition, etc. A full integration of space-sensed spectral information with ground observations, and the generation of accurate predictive models has not been yet successfully achieved in drought-prone ecosystems with complex vegetation structure, large fractions of senescent material and high biodiversity, such as most of Mediterranean TGA. Some methodological questions still require attention and further investigation as for example: 1) how field measurements should be properly used to calibrate and validate models based on optical data, and 2) how to properly use remote sensing sources at different scales (from local to regional and global) to insure accuracy and further applicability of the models.
DiverSpec-TGA will demonstrate that proximal and remote sensors can provide key information, normally unavailable, on functional diversity, biomass growth rate, forage quality and regulatory properties (such as abiotic stresses). This information can assist management decisions and public policies to improve productivity and grassland management and reduce environmental risks at local, regional and national scales. DiverSpec-TGA has the potential to improve parameterization of biophysical processes of vegetation in ecosystem models beyond typical vegetation indices with the availability of meaningful estimations available at significant spatial and temporal scales over TGA ecosytems.
Trustee, training on Remote Sensing for Ecosystem modeling, is an ITN horizon 2020 project that aims to capacitate the next generation of scientists to understand and deal with the increasing pressure of environmental change on ecosystem functioning and land-atmosphere interactions. This project has four main objectives 1) identify essential biodiversity variables (EBVs) and link them with plant traits (PTs) and ecosystem functional properties (EFPs), inferable from remote sensing, 2) investigate new avenues for assessing vegetation photosynthetic efficiency from remote sensing of fluorescence, 3) assimilate diverse remote sensing data streams with varying spatial and temporal resolutions and 4) exploit new satellite missions (e.g. ESA-FLEX, ESA-Sentinels) and earth observation products for the upscaling of PTs, EBVs, and EFPs.
SynerTGE explored sensor synergies for monitoring and modeling key vegetation biophysical variables in tree-grass ecosystems. Mixed tree-grass or shrub-grass vegetation, savanna-like landscapes, are one of the most widely distributed ecosystems on Earth. However, remote sensing and Earth system modeling products are poorly adapted to monitor the key structural and functional characteristics of these ecosystems. SynerTGE proposed to estimate the key vegetation biophysical variables such as leaf area index, canopy water content or leaf pigments using multi-scale sensors to generate products adapted to these tree-grass ecosystems. Novel vegetation indices were developed with the development of three-dimensional radiative transfer models (RTMs) to better account for the structural complexities of the vegetation elements present. These products were then used to improve the estimates of ecosystem scale carbon (e.g. Gross Primary Production) and water (e.g. Evapotranspiration) fluxes.
Fluxpec aimed to improve the monitoring of water and carbon fluxes for a Mediterrean ‘Dehesa’ ecosystem by integrating both proximal and remote sensing. An intensive global effort has been placed on developing novel methods to monitor and model carbon and water exchanges between terrestrial biosphere and atmosphere to better understand the feedbacks of climate change. The project contributed to this by exploring the combination of eddy-covariance continuous measurements with detailed spectral observations, from hyperspectral optical, thermal and LiDAR sensors. Empirical and process-based models were developed and calibrated through extensive and intensive field measurements at the Majadas experimental site, a ‘Dehesa’ Savanna-like ecosystem. The spatial and temporal scaling of these models were investigated through acquisitions from unmanned aerial vehicles (UAV), airborne and satellite sensors.
Biospec aimed at linking multi-scale spectral information with biophysical variables of Mediterranean vegetation in the context of global change. These regions are particularly vulnerable to the effects of the global and climate change. Empirical models were used to derive parameters from the leaf to ecosystem scales. This scaling-up approach was used to quantify uncertainty associated with remote sensing-based estimations of vegetation parameters, which are key indicators for local, regional and global understanding of carbon and water fluxes. Two different Mediterranean case studies were investigated: an Oak savanna-like ecosystem (Spanish dehesa), and a rice paddy.