Three-way analysis of a designed compost experiment using near-infrared spectroscopy and laboratory measurements
Article first published online: 16 FEB 2011
Copyright © 2011 John Wiley & Sons, Ltd.
Journal of Chemometrics
Volume 25, Issue 4, pages 193–200, April 2011
How to Cite
Lillhonga, T. and Geladi, P. (2011), Three-way analysis of a designed compost experiment using near-infrared spectroscopy and laboratory measurements. J. Chemometrics, 25: 193–200. doi: 10.1002/cem.1371
- Issue published online: 14 APR 2011
- Article first published online: 16 FEB 2011
- Manuscript Accepted: 5 NOV 2010
- Manuscript Revised: 24 SEP 2010
- Manuscript Received: 24 MAY 2010
Composts are complex organic systems that undergo batch fermentation processes. Traditional monitoring of such processes is usually based on measuring important chemical (physical) laboratory parameters but the common trend includes using more rapid and non-destructive methods like near-infrared (NIR) spectroscopy. A lab-scale designed (simplex mixture) experiment with nine compost batches, including three repeated centre point batches, was monitored over 5 weeks by NIR spectroscopy (900–1700 nm) and by wet chemical and physical measurements: pH, energy content, moisture content, NH3/NH and temperature. The data were organized in three-way data arrays and different three-way methods were used for analysis: (1) PARAFAC, (2) Tucker3 and (3) PARAFAC2. The present paper stresses the advantages and the possibilities of three-way methods compared to traditional two-way analysis methods such as principal component analysis (PCA). Two-way methods have a tendency to mix variables and produce, from a parsimony point of view, more complex models which are hard to interpret. The results from the three-way methods reproduced the mixture triangle, gave common time profiles (PARAFAC and Tucker3) for all compost batches and rate constants (half-lives) could be calculated: 6.9 days for the PARAFAC loadings from the chemical/physical parameters and between 6 and 10 days for the PARAFAC loadings from the NIR data. PARAFAC2 includes the possibility of getting individual time profiles for each compost batch. The results show that chemical/physical data and the NIR data give similar interpretations. The conclusion is that three-way methods can be used to monitor composts batches over time. Copyright © 2011 John Wiley & Sons, Ltd.