Feasibility Assessment of Aboveground Biomass and Carbon Stock Estimation in Heterogeneous Tropical Campus Green Spaces Using UAV RGB Photogrammetry
More details
Hide details
1
Environmental Science, Universitas Negeri Semarang, Sekaran, Semarang, 50229
2
Departement of Biology, Universitas Negeri Semarang, Sekaran, Semarang, 50229
3
Doctoral Program of Aquatic Resources Management, Faculty of Fisheries and Marine Sciences, Universitas Diponegoro, Semarang, 50275
4
Department of Geography, Universitas Negeri Semarang, Sekaran, Semarang, 50229
5
Department of Biology, National Changhua University of Education, Changhua County, 500
6
PT Muda Karya Geospasial, Demak, 59557
7
Master Program of Urban and Regional Planning, Universitas Diponegoro, Semarang, 50275
8
Center for Sustainable Production System and Life Cycle Assessment, National Research and Innovation Agency (BRIN), Tangerang Selatan, 15314
Corresponding author
Trida Ridho Fariz
Environmental Science, Universitas Negeri Semarang, Sekaran, Semarang, 50229
KEYWORDS
TOPICS
ABSTRACT
Climate change is a global challenge, yet the contribution of higher education institutions to greenhouse gas (GHG) emissions remains insufficiently examined. Campus green open spaces may function as carbon sinks, but biomass and carbon stock estimates in heterogeneous tropical campus environments are still limited. Biomass estimation using Unmanned Aerial Vehicles (UAVs) typically relies on expensive LiDAR sensors, while UAV RGB-based studies are mostly confined to homogeneous forest ecosystems. This study presents a feasibility assessment of using UAV RGB photogrammetry combined with an Individual Tree Canopy (ITC) approach to estimate aboveground biomass (AGB) and carbon stocks in the green open spaces of Universitas Negeri Semarang (UNNES). The workflow includes UAV data acquisition, Ground Control Point (GCP) and Check Point (CP) measurements, orthophoto and Digital Elevation Model (DEM) generation, and Canopy Height Model (CHM) development. Individual tree canopies were delineated through visual interpretation of orthophotos, while diameter at breast height (DBH) data from field surveys were used to calculate reference biomass. AGB models were developed using linear and power regression. The most feasible model was the power regression based on the total CHM values within each canopy, yielding an RMSE of 1770, an MAE of 1348, and a correlation coefficient of 0.41. Although the linear regression model showed slightly better statistical metrics, its raster-scale application produced unrealistic AGB estimates. Spatial aggregation at a 1 × 1 m resolution resulted in a total AGB of 36,962,888 kg for the UNNES campus, corresponding to a carbon stock of approximately 18,481,444 kg C and CO₂ sequestration of 67,822,300 kg CO₂. This study is not intended to replace high-precision LiDAR-based methods, but rather to demonstrate the feasibility of UAV RGB as an estimation approach that is acceptable, stable, and sufficiently replicable in heterogeneous tropical campus contexts, enabling spatially explicit assessments of campus-scale carbon storage.