Field experiment demonstrates the potential utility of satellite-derived reflectance indices for monitoring regeneration of boreal forest communities

Large-scale reforestation efforts require scalable and affordable monitoring technologies to track progress. Boreal forests disturbed by oil and gas development provide an opportunity to utilize ongoing reclamation efforts to assess new monitoring technologies with potential utility to other regions. In this study, we tested the application of reflectance indices derived from high spatial resolution satellite imagery to monitor early-stage reforestation. We installed a randomized nursery experiment in Alberta, Canada wherein the total stocking density and relative abundance of four common boreal tree species were manipulated. Satellite imagery from Maxar's WorldView-2 and WorldView-3 satellites were collected over the experimental plots annually from 2017 through 2020. From these reflectance data, we calculated one texture index, three greenness indices, and one index that integrated both texture and greenness. Finally, two sets of regression models were developed, one of which included only manipulated stem density while the other set included both the density and the mean height of stems in a given year. In the first year of the experiment, we detected no significant differences in any reflectance index across the range of manipulations. Thereafter, all indices showed significant differences, with the integrated texture-greenness index demonstrating the best relative performance among those tested. Our results illustrate the potential for satellite data to yield information regarding tree density and height during the critical early years of restoration. We suggest specific steps that future researchers could use extend this work to other regions and enable potential adoption of this monitoring approach at scale.