Development of automatic spectrometric systems for proximal sensing of photosynthetic activity of vegetation

Terrestrial ecosystems absorb approximately 120 Gt of carbon annually through the photosynthetic process, also referred as Gross Primary Production (GPP). For this reason, assessment and modeling of photosynthetic functioning are critical issues in climate change research and crucial in predicting carbon dynamics. The advent of Eddy Covariance (EC) technique provided(e.g., Baldocchi et al., 1996) in situ long-term and continuous measurements of carbon and water gas fluxes. An extensive network of such EC measurement stations has been established for a better comprehension of the main processes governing the carbon cycle. However, EC technique offers an intensive sampling of the temporal domain, but it lacks in spatial representativeness since it measures a relatively small area. Satellite Remote Sensing (RS) observations provide spatial and temporal variability of ecosystem parameters driving carbon fluxes which can be integrated with EC measurements to scale up carbon estimates to regional and global level. Current Earth observing systems generally provide information of plant status derived from structural or biochemical properties such as Leaf Area Index (LAI) or chlorophyll content. These data can be used to model the potential photosynthetic rates of plant ecosystems and not the actual one. An alternative is offered by the monitoring of the energy dissipation pathways. In fact, the excess light absorbed by vegetation is released through heat dissipation (xanthophyll cycle) and/or re-emitted as Fluorescence (F). Recent RS techniques have encouraged the investigation of these energy dissipation pathways through the Photochemical Reflectance Index (PRI) (heat dissipation) and the analysis of the Fraunhofer lines (F re-emission). For a better comprehension of the link between photosynthesis and optical signals, a growing number of studies have recently focused on the collection of large datasets of repeated field spectral measurements in the sampling area of EC flux towers. Object of this dissertation is the development of automatic spectrometric systems capable of collecting unattended, continuous, long-term spectral measurements of vegetation. Two different automatic spectrometric systems have been developed: the HyperSpectral Irradiometer (HSI) and the Multiplexer Radiometer Irradiometer (MRI). Both instruments are able to routinely and autonomously measure: sun incoming irradiance (downwelling Irradiance, ETOT) and the irradiance/radiance upwelling from the investigated Earth surface (HSI measures ES, MRI measures LS). The systems are able to simultaneously collect “fine” and “ultra-fine” spectrums that allow the computation of different Vegetation Indices (VIs), including the PRI, and the estimation of the sun-induced chlorophyll Fluorescence at O2-A band (F@760). HSI and MRI systems are intended to be operated in experimental sites equipped with Eddy Covariance flux towers to increase our understanding of the link between optical signals and CO2 fluxes. Instruments were installed at two different sites equipped with a micrometeorological EC flux tower. HSI was operated in an alpine pasture for two consecutive years (2009/2010), while MRI was installed in an alfalfa field in the context of the Sen3Exp field survey. Time series of spectral indices (VIs, PRI and F@760) and CO2 fluxes data are reported and analyzed. As a demonstration of the potential of these instruments to estimate the GPP, the use of derived Vegetation Indices and F@760 as inputs of Light Use Efficiency (LUE) models was tested in the last part of the research. Several versions of the basic LUE models were used in order to verify the most effective formulation for the description of GPP.