Measuring post-fire forest regeneration using the greenness index on UAS point clouds to guide restoration in arid forests

With the ongoing climate change, wildfires in arid climates may alter the forest species composition, or even threaten the permanence of forest cover. Land managers face these issues with post-fire forest restoration, often in the form of planting seedlings. Various techniques are available to foster seedlings survival, but monitoring of forest regeneration still relies greatly on field sampling. Regeneration monitoring is necessary to intervene with successive plantation in case of failure, but it is often omitted due to the high costs of field data acquisition. Any method reducing the amount of fieldwork required to estimate postfire forest regeneration would ease the work of land managers in burnt areas. In this case study we tested a methodology to monitor the status of a reforestation in a burnt ponderosa pine forest in Arizona with the use of Uncrewed Aircraft Systems (UAS). A 3D representation of the landscape with point clouds was obtained through Structure From Motion on RGB images. Then, leveraging the colour difference between green pine seedlings and cured vegetation, and the apical dominance of pine seedlings, we were able to identify the individual seedlings in the point clouds and to measure their height. We compared the remote sensing results with actual field monitoring on permanent plots, showing moderate to high accuracy depending on the variable considered. Compared to most monitoring goals, remote sensing offers a highly efficient advantage.