Error propagation in forest biomass assessment.

Forest biomass is the basis for the estimation of carbon storage and emission due to forestry sector. Though the total forest biomass includes aboveground and belowground biomass, this paper deals with the issues related to aboveground biomass. The total aboveground biomass is estimated through a number of variables measured across various components of trees using non-destructive methods. The techniques employed range from simple measuring tape to regression models to satellite imageries. The total error in biomass estimates is the sum of errors in the variables propagated in a hierarchical fashion. The knowledge of prediction errors helps to know the quality of biomass and subsequently bio-energy and carbon estimates. In this paper, various sources of error in biomass estimation, error quantification and error propagation are discussed. The sources of error include tree measurements, sampling strategy, choice of an allometric model and satellite imageries.In South Asia, the standard errors of co-efficient of biomass equations and R 2 are often depicted as indicators for the quality of volumeand biomass equations. The Studies on error propagation in biomass estimates are scarce. Monte Carlo analysis, Pseudo-meta-analysis and Bayesian model averaging have been investigated to address the issues of error propagation in biomass estimation. Among these Bayesian model averaging appears to be a promising technique.