Mapping Relative Health of Individual Spruces with Multispectral Drone

Abstract

Nitrogen is a key nutrient influencing photosynthesis, growth, and resilience in Boreal forest ecosystems. As nitrogen deficiency becomes more prevalent due to anthropogenic pressures, efficient and scalable methods for detecting forest stress are increasingly needed. This study explores the feasibility of using drone-based imaging and spectral data to assess the health status of Picea abies (Norway spruce), with a particular focus on identifying signs of nutrient deficiency and general stress. We proposed and validated three hypotheses: (1) the spectral reflectance of individual spruce needles reveals health status, (2) drone-based 3D mapping can effectively identify and map individual spruce trees, and (3) multispectral drone imagery can detect health differences between trees. Our study was conducted on an 18-hectare spruce-dominant test site in Savonranta, Finland. High-resolution 3D models were created using UAV imagery, allowing for the detection and spatial mapping of over 3,000 individual spruce trees. Needle-level spectral measurements showed clear reflectance differences between healthy trees, nitrogen-deficient trees, and those under additional stress. Multispectral drone data further supported these findings, with particularly strong differentiation observed in the red edge and near-infrared bands. These results demonstrate that drone-based remote sensing techniques can detect subtle physiological changes in trees and offer a scalable approach to forest health monitoring. The integration of in-situ measurements with high-resolution aerial data presents a promising direction for sustainable forest management in the context of global environmental change.