Search

We propose a combination of portable X-ray fluorescence (pXRF) and chemometrics to discriminate between Mesoamerican obsidian sources and to assign archaeological artifacts of unknown origin to their respective deposits using a procedure that does not require any type of calibration or reference standards. A set of 109 samples of known origin and a total of 257 samples of unknown origin were analyzed with a portable XRF spectrometer. The resultant spectra were used as spectral signatures for the chemometric data analysis. First, we applied spectral pre-treatment techniques, such as CluPA algorithm for peak alignment and the Savitzky-Golay and Extended Multiplicative Signal Correction for data smoothing and noise removal, combined with methods for the selection of a spectral range containing the variables with the most relevant information (iPLS and). The full spectrum of the obsidian samples was divided into 20 subintervals, fitting a local regression model (PLS) to each subinterval. The performance was evaluated by the Root Mean Square Error of Cross-Validation, the Root Mean Square Error of Prediction and the correlation coefficient. The selected spectral regions were then analyzed with ROBPCA algorithm for the discrimination of outliers and the projection of the observations in the PCA space. For the classification, we propose a robust procedure (RSIMCA) which is based on a ROBPCA method for high-dimensional data. The classification rules were obtained by using the orthogonal and the score distances, from which it is possible to distinguish samples that belong to a given group. Using the proposed methodology, we were able to provide evidence about which variables were meaningful for the classification and provided information about group membership or provenance. This approach proves to be a valid technique for the quantitative analysis of XRF spectra. [ABSTRACT FROM AUTHOR]

Abstract: A growing number of investigations have shown that remote sensing of foliar nitrogen (N) concentration in plant canopies can be achieved with imaging spectroscopy, or hyperspectral remote sensing, from satellite or airborne sensors. Development of this approach has been fueled by recognition that foliar N is related to a variety of ecological and biogeochemical processes, ranging from the spread of invasive species to the ecosystem effects of insect defoliation events to patterns of N cycling in forest soils. To date, most studies have focused on building site-specific foliar N detection algorithms applied to individual scenes or small landscapes that have been intensively characterized with local field measurements. However, the growing number of well-measured sites, combined with improvements in image data quality and processing methods provide an opportunity to begin seeking more general N detection methods that can be applied to a broader range of sites or to locations that lack intensive field measurements. Here, we combine data from several independent efforts in North America, Central America and Australia, to examine whether development of calibration methods to determine canopy nitrogen concentration across a wide range of forest ecosystems is possible. The analysis included data from 137 individual field plots within eight study sites for which imagery has been acquired from NASA''s Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) and/or Hyperion instruments. The combined dataset was used to evaluate site-specific calibration results as well as results obtained with data pooled across all sites. We evaluated the accuracy of results using plot- and site-level cross-validation wherein individual plots or entire sites were withheld and used as an independent validation of the resulting algorithms. In instances where all sites were represented in the calibration, canopy-level foliar N concentration was predicted to within 7–15% of the mean field-measured values indicating a strong potential for broadly applied foliar N detection. When whole sites were iteratively dropped from the calibration and predicted by remaining data, predictions were still significant, but less accurate (7–47% of mean canopy-level N concentration). This suggests that further development to include a wider range of ecosystems will be necessary before cross-site prediction accuracy approaches that seen in site-specific calibrations. Nevertheless, we view these results as promising, particularly given the potential value of foliar N estimates, even at a reduced level of confidence, at sites for which there is no possibility of conducting field data collections. [Copyright &y& Elsevier]