Testing the utility of multivariate techniques in mapping commercial forest species using freely available Landsat 8 Operational Land Imager (OLI)

ABSTRACT

Reliable species discrimination remains essential for the management of commercial forests. Therefore, this study sought to evaluate the utility of Partial Least Squares Linear Discriminant Analysis (PLS-LDA) and Partial Least Squares-Discriminant Analysis (PLS-DA) multivariate techniques for delineating forest species using Landsat 8 OLI. PLS-LDA produced a higher (88.9%) overall accuracy compared to the PLS-DA (79%). The high performance of PLS-LDA is associated with its ability to deal with correlation and variability between and within classes, hence offer great potential for the monitoring and management of commercial forest species.

KEYWORDS:

• Forest species
• multivariate techniques
• classification

Acknowledgments

The authors send gratitude to the University of KwaZulu-Natal and NRF program for granting this research opportunity and financial support. We also thank the anonymous reviewers for their constructive input.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental material for this article can be accessed here.

Additional information

Funding

This work was supported by the National Research Foundation [107232 & 119791].