2009: In-Field KPY
Field-based measures of Kraft pulp yield
Between 2001 and 2009 a series of FWPA-funded projects were initiated to determine whether portable NIR spectrometers can be used to measure Kraft Pulp Yield, Dry Matter Content and Basic Density using spectra taken in the field from non-destructive samples.
-
Multi-site NIR calibrations were generated for Kraft pulp yield, basic density and dry matter content. These combined 120 E. globulus and 80 E. nitens trees from 40 contrasting sites across Tasmania, Victoria, South Australia and Western Australia into a single calibration model.
-
Separating samples into species or regional calibrations did not improve the relationships.
-
The three wood properties can be predicted from spectra acquired from standing trees using portable NIR instruments. Calibrations for KPY were not as strong as had been expected, attributable in large part to within-tree variation between the properties of the breast height sample from which spectra were collected and the properties of the whole-tree sample from which KPY data were determined.
-
The hand-held Polychromix Phazir instrument was found to be a practical and robust field instrument well suited to this application. Recent developments in the instrument may make its wavelength range more suitable.
-
Calibrations developed using spectra acquired on one Phazir can be transferred to other instruments of the same type.
-
Chainsaw dust was found to be the best sample type, yielding the strongest calibration models. Increment cores (green and dry) also yielded useful relationships.
-
Independent test set: Calibrations based on spectra acquired from green chainsaw frit on the laboratory-based NIR instrument provided accurate and precise predictions of basic density and KPY, of an independent set of 50 trees sampled by Oji Plantations and pulped by Gunns (basic density r2 = 0.80; KPY r2 = 0.73). Given that increment core density itself only explained 85% of whole-tree chip density, the 80% explained by the NIR predictions was better than expected. Predictions based on Phazir spectra of samples explained 70% of the variance in density and KPY (Appendix G).
-
Acoustic Wave Velocity (AWV) from standing trees explained no more than 30% of the variance in KPY. Correlations between AWV and chip basic density were weaker, and non-existent with dry matter content.