The best practice of forest inventory using remotely -sensed data in tropical forest, Indonesia

Abstract

This paper described the implementation of the conventional terrestrial-based comprehensive and periodic forest inventory (TBFI) and the image-based forest inventory (IBFI) in nine forest management units, in the tropical forest of Indonesia. The study examined the best practices of high-resolution and very high-resolution imagery-based forest inventory. The aspects studied include 1) cost comparison, 2) completion time, and 3) sampling error. This study found that the IBFI was more efficient than TBFI, having a cost ratio from 0.358 to 0.859. Technically, the IBFI is more implementable than TBFI. The cost per unit area decreases as the area to be surveyed increases. The average cost per unit area of IBFI is always lower than the TBFI. From the duration of the survey and data analysis, the IBFI method could be completed from 42 days to 83 days, much faster than the TBFI method. The study also found the relative efficiency of image-based double sampling much more efficient than terrestrial simple random sampling.