É«É«À²

The title of the thesis: Characterizing Boreal Forest Structure and Estimating Above-Ground Biomass with Helicopter-borne Ku-band Profiling Radar

Thesis defender: Ziyi Feng
Opponent: Prof. Johan Fransson, Linnaeus University, Sweden
Custos: Prof. Jaan Praks, Aalto University School of Electrical Engineering

Short-wavelength Ku-band radar signals can penetrate boreal forest canopies more effectively than commonly assumed. This doctoral research shows that a lightweight airborne radar can identify the ground and canopy top and provide information for estimating tree height, stem volume, and above-ground biomass. Reliable measurements of these attributes are important for climate research, forest monitoring, and the sustainable use of renewable forest resources.

Radar measurements can be made day or night and through cloud cover. However, Ku-band signals have often been assumed to reflect mainly from the upper canopy, limiting their value for describing the whole forest profile. The research tested this assumption using Tomoradar, a lightweight system developed by the Finnish Geospatial Research Institute and mounted on a helicopter over the Evo test area in southern Finland. Unlike a conventional radar image, each measurement records the strength of radio echoes from different heights below the helicopter, producing a vertical profile of the forest. The radar data were compared with simultaneous airborne laser scanning data and field measurements.

Under stable downward-looking flight conditions, the radar identified the ground with an average error of approximately 0.4 meters. It identified the canopy top with an average error of approximately 0.8 meters. A new radar indicator describing how reflected energy was distributed between the canopy and the full forest profile correlated strongly with a laser-based measure of canopy density. At sample-plot level, radar data were also used to estimate mean tree height, mean trunk diameter, stem volume, and above-ground biomass. The relative errors ranged from about 10% for mean height to about 20% for stem volume and biomass.

The findings improve understanding of how short-wavelength radar signals interact with boreal forests and demonstrate that Ku-band profiling radar can complement laser scanning and satellite radar observations. With further development for drones and integration with other sensors, the method could support flexible and cost-efficient forest inventories and better monitoring of forest carbon stocks.

Key words: Ku-band radar; profiling radar; boreal forest; forest structure; above-ground biomass; remote sensing

Thesis available for public display 7 days prior to the defence at .