A hybrid approach of remote sensing for mapping vegetation biodiversity in a tropical rainforest

Abstract

Abstract. Wardhana W, Widyatmanti W, Soraya E, Soeprijadi D, Larasati B, Umarhadi DA, Hutomo YHT, Idris F, Wirabuana PYAP. 2020. A hybrid approach of remote sensing for mapping vegetation biodiversity in a tropical rainforest. Biodiversitas 21: 3946-3953.  Vegetation biodiversity is one of the most important indicators to evaluate the sustainability of tropical rainforest. It is commonly described by three essential variables, i.e. richness, heterogeneity, and evenness. That information is frequently collected from periodic forest inventory using terrestrial method. However, this effort needs a long-time consuming, high cost, and almost impossible to implement in the area of tropical rainforest with hard accessibility. This study investigates the potential of remote sensing as an alternative method for mapping vegetation biodiversity in a tropical rainforest. A hybrid approach of remote sensing using medium and high-resolution images was developed to recognize the attributes of vegetation biodiversity by considering three parameters derived from remote sensing data, including canopy density (C), crown diameter (D), and tree density (N). The use of a medium resolution image aimed to categorize vegetation density using Modified Soil-Adjusted Value Index (MSAVI) while a high-resolution image was utilized to acquire a more detailed spectrum for determining C, D, and N in every class of vegetation density. The relationship between C, D, N, and richness, heterogeneity, evenness was explained using hierarchical cluster analysis. Our study discovered the attributes of vegetation biodiversity in a tropical rainforest could be potentially recognized by combining C, D, and N as predictor variables.