The Interpolation Method for Estimating the Above-Ground Biomass Using Terrestrial-Based Inventory

Abstract

This paper examined several methods for interpolating biomass on logged-over dry land forest using terrestrial-based forest inventory in Labanan, East Kalimantan and Lamandau, Kota Wringing Barat, Central Kalimantan.  The plot-distances examined was 1,000−1,050 m for Labanan and 1,000−899m for Lawanda.  The main objective of this study was to obtain the best interpolation method having the most accurate prediction on spatial distribution of forest biomass for dry land forest. Two main interpolation methods were examined: (1) deterministic approach using the IDW method and (2) geo-statistics approach  using Kriging with spherical, circular, linear, exponential, and Gaussian models.   The study results at both sites consistently showed that the IDW method was better than the Kriging method for estimating the spatial distribution of biomass.  The validation results using chi-square test showed that the IDW interpolation provided accurate biomass estimation.   Using the percentage of mean deviation value (MD(%)), it was also recognized that the IDWs with power parameter (p) of 2 provided relatively low value , i.e., only 15% for Labanan, East Kalimantan Province and 17% for Lamandau, Kota Wringing Barat Central Kalimantan Province. In general, IDW interpolation method provided better results than the Kriging, where the Kriging method provided MD(%) of about 27% and 21% for Lamandau and Labanan sites, respectively.