Estimation of Forest Structural Parameter using Remote Sensing Technology in Central Mindanao University

Abstract

Effective monitoring of natural forest ecosystems requires efficient and scalable approaches to address the limitations of conventional field-based measurements, which are often labor-intensive, costly, and spatially constrained. This study explores the application of Sentinel-2 multispectral imagery for assessing forest structural parameters in the natural forest of Central Mindanao University (CMU), Bukidnon, Philippines. Field data on crown length, tree frequency, and basal area were collected from fifteen sample plots and compared with remote sensing–derived vegetation indices, including the Normalized Burn Ratio (NBR), Moisture Vegetation Index (MVI), and Sentinel-2 Band 2 reflectance. Statistical analyses revealed a strong correlation between crown length and the combined indices of NBR, MVI, and Band 2 reflectance, with an adjusted R² of 0.885, highlighting their capability to capture canopy moisture status, disturbance intensity, and understory conditions. In contrast, tree frequency showed a moderate relationship with maximum NBR values (adjusted R² = 0.339), suggesting that individual indices have limited explanatory power for certain structural attributes. Spatial analysis further demonstrated that undisturbed forest core areas exhibit longer crown lengths, while fragmented and peripheral zones are characterized by shorter crowns, reflecting the impacts of human activities and subsequent forest regeneration. Overall, the results indicate that Sentinel-2 imagery provides a cost-effective and scalable framework for forest condition assessment, supporting adaptive forest management, conservation of mature forest patches, and informed planning for reforestation and assisted natural regeneration in disturbed areas.