Table 1 Description of merits and demerits of methods for above-ground biomass estimation in forest ecosystems
From: Estimation of above-ground biomass in tropical afro-montane forest using Sentinel-2 derived indices
Above-ground biomass estimation methods | ||
|---|---|---|
Direct measurement (destructive method) | Indirect measurement (non-destructive method) | |
Biomass equations (allometric equations) | Coupled methods (remote sensing and statistics) | |
• Involves direct harvesting of all trees/shrubs • Biomass measured from oven dry mass of tree components (stem, braches, twigs, leaves) Advantages: • High accuracy • Reliable data/information • No need for validation • Useful for developing species-specific biomass equation Disadvantages: • Only feasible/applied for small area and for small number of trees • Costly and time consuming • Not applicable for rare or threatened species • Not done for inaccessible terrains (Brown et al. 1989; Brown 1993; MoA 2000; Zianis and Mencuccini 2004; Shrestha 2011; Vashum and Jayakumar 2012; Yohannes et al. 2015) | • Species-specific or mixed-species regression equations developed from felled sample trees • Biomass estimated from measured tree parameters Advantages: • No harvesting of trees, except for few samples • Useful for protected forests with rare and threatened species • Cost and time efficient Disadvantages: • Low accuracy/reliable • Results may be under- or over estimated • Field inventory/tree parameters needed • Less useful for inaccessible terrains • Accounts only for live trees/shrubs • Requires validation by felling trees • Site specific (e.g., rainfall) (Brown 1997; Pearson et al. 2005; Yohannes et al. 2015; Siraj 2019; Dibaba et al. 2019) | • Remote sensing measures vegetation attributes that are correlated with biomass • Biomass prediction models developed using field data • Biomass estimated from remotely sensed data using a prediction model Advantages: • Captures remotely sensed vegetation attributes • Useful for inaccessible terrain, large areas at scale • Reliable and cost effective • Useful for monitoring vegetation • Model is developed from sample measurements • Uses spectral band width, vegetation index, variables • Several options of sensors (Landsat, SPOT, MODIS, LiDAR, Sentinel, etc.…) Disadvantages: • Low availability of images • High cost for high resolution • Limitations in resolutions • Limited comparability across sensors • Objective specific/thematic • Complexity of classification (Lyon et al. 1998; Lu et al. 2014; Timothy et al. 2015; Georgia et al. 2017; Castillo et al. 2017; Pertille et al. 2019; Isbaex and Coelho 2020) |