Open Collections

Abstract

Recent wildfires have burned millions of hectares of North American forests, including over 4% of British Columbia since 2017. Post-fire monitoring in British Columbia lacks spatial and temporal continuity, limiting our capacity to predict and respond to changes in forest composition. Earth Observations satellites, such as Landsat, can help fulfil these data gaps; however, long-term observations require validating post-fire spectral responses to changes in recovering forest structure. Aligning the spatial and temporal scales of satellite imagery with field observations is challenging. Complementary remote sensing technologies, such as light detection and ranging (lidar) data from Remotely Piloted Aircraft (RPAs), can complement field observations and help estimate forest structure at scales that align with satellite data. This dissertation links RPA lidar estimates of conifer recovery to unique spectral responses from Landsat, highlighting the importance of integrating different remote sensing technologies in monitoring. First, RPA estimates of conifer structure were related to spectral time-series data 16 years post-fire. Unique spectral trends were associated with distinct forest composition and density. Building on these findings, early spectral trends (as early as five years after fire) were matched to rates and patterns of conifer establishment, based on a space-for-time RPA lidar dataset 5-20 years post-fire. Research identified two groupings of post-fire recovery across interior BC. The two groupings represented mixed re-growth or early deciduous growth, followed by dense conifer growth. To investigate controls on conifer recovery, structural groups were associated with ecological drivers, such as post-fire drought or pre-fire basal area. Structural recovery groupings were best predicted by post-fire weather and pre-fire composition. Lastly, the structural groupings and associated ecological drivers were systematically compared to broader research on post-fire conifer recovery. Importantly, the recovery observed in interior BC aligned with conifer recovery patterns and trends observed elsewhere in western North America. Temporal spectral trends from satellites played a crucial role in distinguishing the timing of conifer recovery, including delayed conifer establishment. Overall, this research demonstrates that combining the historical archive of Landsat imagery with fine-scale forest structure derived from RPAs lidar enables spatially and temporally monitoring of post-fire recovery.