Influence of Plot Size on Efficiency of Biomass Estimates in Inventories of Dry Tropical Forests Assisted by Photogrammetric Data from an Unmanned Aircraft System

• Main Authors: Daud Jones Kachamba, Hans Ole Ørka, Erik Næsset, Tron Eid, Terje Gobakken
• Format: Article
• Language: English
• Published: MDPI AG 2017-06-01
• Series: Remote Sensing
• Subjects: miombo woodlands; plot size; model-assisted inferential framework; relative efficiency
• Online Access: http://www.mdpi.com/2072-4292/9/6/610

Applications of unmanned aircraft systems (UASs) to assist in forest inventories have provided promising results in biomass estimation for different forest types. Recent studies demonstrating use of different types of remotely sensed data to assist in biomass estimation have shown that accuracy and precision of estimates are influenced by the size of field sample plots used to obtain reference values for biomass. The objective of this case study was to assess the influence of sample plot size on efficiency of UAS-assisted biomass estimates in the dry tropical miombo woodlands of Malawi. The results of a design-based field sample inventory assisted by three-dimensional point clouds obtained from aerial imagery acquired with a UAS showed that the root mean square errors as well as the standard error estimates of mean biomass decreased as sample plot sizes increased. Furthermore, relative efficiency values over different sample plot sizes were above 1.0 in a design-based and model-assisted inferential framework, indicating that UAS-assisted inventories were more efficient than purely field-based inventories. The results on relative costs for UAS-assisted and pure field-based sample plot inventories revealed that there is a trade-off between inventory costs and required precision. For example, in our study if a standard error of less than approximately 3 Mg ha−1 was targeted, then a UAS-assisted forest inventory should be applied to ensure more cost effective and precise estimates. Future studies should therefore focus on finding optimum plot sizes for particular applications, like for example in projects under the Reducing Emissions from Deforestation and Forest Degradation, plus forest conservation, sustainable management of forest and enhancement of carbon stocks (REDD+) mechanism with different geographical scales.