A Lidar-Radar Framework to Assess the Impact of Vertical Forest Structure on Interferometric Coherence

In this paper, we present novel modeling approaches to investigate the sensitivity of radar interferometric coherence to variations in the vertical forest canopy profile. We introduce a common framework applicable to model radar microwave extinction and structure from lidar data. To perform this analysis, we make use of interferometric data from the uninhabited aerial vehicle synthetic aperture radar (UAVSAR) L-band radar and full waveform lidar data from laser vegetation imaging sensor (LVIS). The datasets were acquired over the Laurentides Wildlife Reserve Forest, Quebec, Canada. A twofold analysis of the framework to estimate interferometric coherence is undertaken. First, a sensitivity analysis is performed by incorporating lidar waveform Legendre descriptions into two adapted independent polarimetric interferometry models. Second, we examine the effectiveness of using lidar data in this novel way to model radar interferometric coherence. Where appropriate, coherence estimates are obtained using Legendre solutions up to fourth order and at resolutions up to 75 m. The maximum r2 values between modeled outputs and observed coherence across hh, vv, and hv polarizations are shown as 0.51(p < 0.05) and 0.76(p < 0.05) at 25 and 75 m pixel resolutions, respectively. The introduction of a common framework to combine lidar and radar enables an estimation of the impact of canopy structure on observed interferometric coherence and provides further insight into the feasibility of assuming uniform microwave extinction rates on different scales through forest canopy. The framework’s potential lies in its use to assess performance of canopy structure estimates from future spaceborne radar interferometers in synergy with lidar data.