Data: Estimating Aboveground Biomass Using Allometric Models and Adaptive Learning Rate Optimization Algorithm

Published: 12-03-2021| Version 1 | DOI: 10.17632/j99jjpbhrv.1


Forest inventory was conducted in 50 plots of 30 m x 30 m sample plots randomly laid in the forest reserve. The structural variables such as diameter at breast height (DBH) ≥ 20 cm, tree height and wood density were recorded in the field. The structural information obtained was used to estimate the AGB, which is the total amount of living organic material of trees. Field measurement of tree variables were carried out using relascope, haga altimeter, increment borer, scale weight, measuring tape, ranging pole and Global Positioning System (GPS). Four allometric models were used with two optimization algorithms; Modified Root Mean Square Propagation (Modified RMSProp) and Modified Adaptive Moment Estimation (Modified Adam) were also used to train each model. This study seeks to evaluate the adaptive learning rate optimization algorithms on allometric model and to determine its efficiency in calculating and predicting above ground biomass with the main purpose of coming up with allometric equations for estimating aboveground biomass in for tropical regions.