Estimation of aboveground biomass of a production forest reserve in Malaysian Borneo using K-nearest neighbor method
This study examined the use of the k-nearest neighbour (k-NN) method to estimate aboveground biomass of a logged-over tropical forest in Sabah, Malaysia. To estimate aboveground biomass, field data as well as digital number and normalised difference vegetation index (NDVI) values from Landsat TM-...
Saved in:
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Forest Research Institute Malaysia
2014
|
Subjects: | |
Online Access: | https://eprints.ums.edu.my/id/eprint/19605/1/Estimation%20of%20aboveground%20biomass.pdf https://eprints.ums.edu.my/id/eprint/19605/ https://www.researchgate.net/publication/265510712_Determining_aboveground_biomass_of_a_forest_reserve_in_Malaysian_Borneo_using_k-nearest_neighbor_method_Journal_of_Tropical_Forest_Science |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Universiti Malaysia Sabah |
Language: | English |
Summary: | This study examined the use of the
k-nearest neighbour (k-NN) method to estimate aboveground biomass of a logged-over tropical forest in Sabah,
Malaysia. To estimate aboveground biomass, field data as well as digital number and normalised difference
vegetation index (NDVI) values from Landsat TM-5 data were used to determine the optimum horizontal
reference area and the number of reference sample plots (k). An accuracy assessment showed that enhancing the
digital number value was superior to enhancing the NDVI value. Root mean square errors of no filtering and 3 × 3
filtering were minimum when k = 4 and k = 5 respectively, when a horizontal reference area of 17 km was applied.
The bias was underestimated by 2.01 and 1.62 tonnes ha-1 for k = 4 and k = 5 respectively. Total aboveground
biomass of the forest management unit estimated by enhancing the digital number value was 6,873,299 tonnes
and average biomass density was 248.8 tonnes ha-1. The results suggest that the k-NN method is an alternative way
to estimate and map aboveground biomass of a forest management unit. |
---|