Radar Measurements of Temporal Variation in a Boreal Forest

Licentiatavhandling, 2018

Synthetic aperture radar (SAR) on a satellite platform is a suitable technique for all-weather global monitoring of forest parameters such as biomass. This is important for increasing the accuracy of the global terrestrial carbon flux, the largest uncertainty in our current understanding of the Earth’s carbon cycle. In recent decades there has been increasing interest in relating SAR observables from current and future spaceborne SAR missions to forest parameters. Ideally, these SAR observables are only dependent on the forest parameters of interest, but they are also affected by temporal variations in the forest due to weather, diurnal and seasonal changes. These effects are not properly accounted for in current forest parameter estimation models using SAR data due to shortcomings in our understanding of these effects. To fill this knowledge gap, a tower-based radar has been developed for measuring temporal variation of radar signatures in a boreal forest site in southern Sweden. The instrument allows monitoring of radar signatures from the forest site over timescales ranging from less than a second to years. The experiment consists of a 50-m high tower equipped with 30 antennas, allowing fully-polarimetric tomographic imaging at microwave frequencies of P-band (420 - 450 MHz), L-band (1240 - 1375 MHz) and C-band (5250 - 5570 MHz). Results from an on-site weather station assists in interpretation of the radar results. The work in this thesis involves the design and implementation of the experimental setup and analysis of the first results. The results have provided new information about variations in forest radar backscatter during freeze-thaw cycles at P- and L-band, causing drops in backscatter of 4 to 10 dB. An effect where strong winds cause a drop in co-polarised backscatter at P-band has been directly observed for the first time. Finally, the tomographic capabilities of the instrument were demonstrated. Tomographic imaging quality was shown to be better than for measurement protocols used by previous tower-based radars when acquired while the trees were moving due to wind. The results of this experiment are important for understanding and modelling temporal variations in radar measurements over boreal forests such that these unwanted variations can be compensated for in forest parameter estimation algorithms using SAR data.