No abstract

No abstract

Abstract. Canopy nitrogen (N) concentration and content are linked to several vegetation processes. Therefore, canopy N concentration is a state variable in global vegetation models with coupled carbon (C) and N cycles. While there are ample C data available to constrain the models, widespread N data are lacking. Remotely sensed vegetation indices have been used to detect canopy N concentration and canopy N content at the local scale in grasslands and forests. Vegetation indices could be a valuable tool to detect canopy N concentration and canopy N content at larger scale. In this paper, we conducted a regional case-study analysis to investigate the relationship between the Medium Resolution Imaging Spectrometer (MERIS) Terrestrial Chlorophyll Index (MTCI) time series from European Space Agency (ESA) Envisat satellite at 1 km spatial resolution and both canopy N concentration (%N) and canopy N content (N g m−2, of ground area) from a Mediterranean forest inventory in the region of Catalonia, in the northeast of Spain. The relationships between the datasets were studied after resampling both datasets to lower spatial resolutions (20, 15, 10 and 5 km) and at the original spatial resolution of 1 km. The results at higher spatial resolution (1 km) yielded significant log–linear relationships between MTCI and both canopy N concentration and content: r2 = 0.32 and r2 = 0.17, respectively. We also investigated these relationships per plant functional type. While the relationship between MTCI and canopy N concentration was strongest for deciduous broadleaf and mixed plots (r2 = 0.24 and r2 = 0.44, respectively), the relationship between MTCI and canopy N content was strongest for evergreen needleleaf trees (r2 = 0.19). At the species level, canopy N concentration was strongly related to MTCI for European beech plots (r2 = 0.69). These results present a new perspective on the application of MTCI time series for canopy N detection.

) from a Mediterranean forests inventory in the region of Catalonia, NE of Spain. The relationships between the datasets were studied after resampling both datasets to lower spatial resolutions (20 km, 15 km, 10 km and 5 km) and at the initial higher spatial resolution of 1 km. The results at the higher spatial resolution yielded significant relationships between MTCI and both canopy N concentration and content, r 2 = 0.32 and r 2 = 0.17, respectively.We also investigated these relationships per plant functional type. While the relationship between MTCI and canopy N 20 concentration was strongest for deciduous broadleaf and mixed plots (r 2 = 0.25 and r 2 = 0.47, respectively), the relationship between MTCI and canopy N content was strongest for evergreen needleleaf trees (r 2 = 0.20). At the species level, canopy N concentration was strongly related to MTCI for European Beech plots (r 2 = 0.71). These results present a new perspective on the application of MTCI time series for canopy N detection, ultimately leading towards the generation of canopy N maps that can be used to constrain global vegetation models. capacity (Evans, 1989;Reich et al., 1995;Reich et al., 1997;Reich et al., 1999;Wright et al., 2004), specific leaf area, leaf life span (Reich et al., 1999;Wright et al., 2004) and light use efficiency (Kergoat et al., 2008). Leaf N concentration expressed on a leaf area basis, also called leaf N content (N g m -2 ) has also been linked with chlorophyll content, Rubisco content (Evans, 35 1989) and photosynthetic capacity (Evans, 1989;Reich et al., 1995). At stand scale, canopy nitrogen concentration, which represents the leaf N concentration averaged over the stand canopy, has also been found to correlate with above ground Net Primary Productivity (NPP) (Reich, 2012), while canopy N content has been linked with the canopy light use efficiency (Green et al., 2003).Given their links to many vegetation processes, leaf and canopy N variables could be used to constrain N cycle modules in 40 global vegetation models. At the global scale, ample data is available to constrain models for the C cycle; however, data to constrain the N cycle are limited. Currently, canopy N data is not widely available and canopy N sampling campaigns are timeconsuming and thus expensive tasks. Moreover, upscaling from local sampling campaign measurements represents an additional limitation. In this perspective, local, regional or even global remotely sensed canopy N estimates will be a valuable addition, enabling us to collect information in a less time intensive and expensive manner than traditional on-field sampling 45 campaigns. Such near global canopy N estimates will be beneficial as input in global vegetation models or to calibrate and validate these models.Currently, different remote sensing techniques have been applied to detect canopy N in terrestrial vegetation. Imaging spectrometry from either airborne or spaceborne sensors coupled with different analysis methods, including partial least squares regression (PLS), continuu...

Chapter 1 Introduction 1.1 Relevance 1.2 The carbon cycle and the terrestrial biosphere 1.3 N cycle and canopy N 1.4 Vegetation models 1.5 Canopy N remote sensing 1.6 Research questions and outline of the thesis 1.7 References Chapter 2 Exploring the use of vegetation indices to sense canopy nitrogen to phosphorous ratio in grasses 2.1 Introduction 2.2 Material and methods 2.2.1 Culture of the plants 2.2.2 Reflectance measurements 2.2.3 Leaf chemical measurements 2.2.4 Spectral bands considered: original and resampled to satellite bands 2.2.5 Data analysis 2.3 Results 2.3.1 Descriptive statistics of canopy N:P, canopy N and canopy P 2.3.2 Original narrow band spectra 2.3.3 Spectra resampled to satellite sensors' bands 2.4 Discussion 2.4.1 Original narrow band spectra Canopy nitrogen: a remote sensing and modelling approach