Potential ecological impacts of disturbance, land use, and climate change have driven many studies to evaluate ecosystem functions through the measurement of vegetation biochemical properties that provide integral information on nutrient cycling, litter decomposition, and plant productivity. The use of spectroscopy in quantifying vegetation biochemistry shows promise with faster analytical speed than traditional methods. Synergies between the Visible Near Infrared/ Short Wave Infrared (VSWIR) and Thermal Infrared (TIR) spectra for identifying plant species' foliar chemistry have been largely unexplored. Here we evaluate the capability of VSWIR and/or TIR spectra to predict leaf levels of lignin, cellulose, nitrogen, water content, and leaf mass per area. We specifically examined how these predictive relationships might change seasonally and among plant functional types. Lastly we determined whether these relationships between spectra and foliar chemistry could be extended to the reduced spectral resolution available in airborne sensors, including the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS), the Hyperspectral Thermal Emission Spectrometer (HyTES), and the combined AVIRIS and MODIS/ASTER (MASTER) sensors used in the Hyperspectral Infrared Imager (HyspIRI) preparatory flight campaign.
In the 2013 spring, summer, and fall seasons, fresh leaves from sixteen common shrub and tree species in California representing three broad plant functional types were sampled from the Sierra Nevada Mountains, the Central Valley at the Sedgwick Reserve, and coastal Santa Barbara. Partial least squares regression (PLSR) analysis was used to relate spectral response at wavelengths from 0.3 - 15.4 µm to laboratory-measured biochemical and biophysical properties. For each component, three PLSR models were fit using different portions of the spectrum: VSWIR (0.3 - 2.5 µm), TIR (2.5 - 15.4 µm), and the full spectrum (0.3 - 15.4 µm). Three additional models were fitted using spectra resampled to AVIRIS (0.4 - 2.5 µm), HyTES (7.5 - 12 µm), and the combined AVIRIS and MASTER (0.38 - 12 µm).
The majority of the highest performing laboratory spectra models used either the TIR or full spectrum. When using simulated sensor spectra, the combined AVIRIS and MASTER produced the highest performing models, followed by HyTES. From both laboratory and sensor simulated model results, the combination of VSWIR and TIR increased the R2 value of regression models compared to VSWIR alone, signifying that the inclusion of TIR data would improve predictions of foliar chemistry. We also found that model precision varied by seasons and across plant functional types. Models developed for all seasons resulted in a decreased R2 value, but still had high precision (R2 > 0.85) and accuracy (RMSE < 10%) when predicting cellulose, nitrogen, and water content. These results indicate that the TIR could augment the VSWIR in advancing identification of leaf properties of the world's ecosystems by helping to set the foundation for future use of the full spectrum represented by the proposed HyspIRI space-borne sensor.