Date of Award
Summer 8-4-2022
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Geography
First Advisor
Wenge Ni-Meister
Second Advisor
Andrew Reinmann
Academic Program Adviser
Sean Ahearn
Abstract
Our knowledge of the distribution and amount of terrestrial above ground biomass (AGB) has increased using lidar technology. Recent advancements in satellite lidar has enabled global mapping of forest biomass and structure. However, there are large biases in satellite lidar estimates which impacts our understanding of carbon dynamics, particularly in tropical forests.
Ni-Meister et al. (2022) developed a lidar full waveform weighted height-based allometric model which produced very good results in temperate deciduous/conifer forest in the continental US. The purpose of this study was to evaluate this biomass model in an African tropical forest using the Land Vegetation and Ice Sensor (LVIS) lidar system. The results were compared with field measured AGB derived from a generalized pan-topical AGB equation (Chave et al. 2014).
Our analysis shows that the biomass model outperforms two regression based biomass models using LVIS and small footprint lidar data. It performs very well (R2=0.84, RMSE=55.67), producing similar results to the best fitted RH empirical model (R2=0.87, RMSE=49.02). However, the biomass model outperforms the RH model when including the wood density parameter from field data (R2=0.91, RMSE=40.47). The height scaling exponent estimated using site-based allometric relationships from individual tree structure and literature data matches well with the optimal height scaling exponent through fitting the model prediction and field data. Testing in a disturbed/young forest site indicates a slight larger scaling exponent and provide much more accurate AGB estimates for young stands. This result implies that the allometric relationships might be different for young and mature forest stands even for the same forest species. The larger scaling exponent for young stands than mature stands also suggests strong AGBD and height dependence for young stands than mature stands. Our model captures the nature of AGBD dependence on height and crown size structure features. The large returns shown in waveforms for mature trees suggests large dependence ABGD on crown size properties for mature forest stands. Our assessment results that this biomass model can be expanded to estimate AGB density in tropical forest biomes using the GEDI satellite lidar data with good accuracies.
Recommended Citation
Rojas, Alejandro, "Quantifying Aboveground Biomass in a Tropical Forest Using a Lidar Waveform Weighted Allometric Model" (2022). CUNY Academic Works.
https://academicworks.cuny.edu/hc_sas_etds/936