Assessing the impact of afforestation on forest vegetation structure using fraction of vegetation cover and leaf area index indicators from harmonized Landsat and Sentinel-2 imagery

Reliable regional monitoring of afforested sites is hampered by the cost of in-situ surveys and limitations of land-cover mapping for assessing canopy condition. The SL2P-CCRS neural network algorithms were applied to harmonized Landsat and Sentinel-2 imagery to derive annual maximum leaf area index (LAI) and fraction of vegetation cover (FCOVER) for 994 afforested sites across a 168,241 km 2 region in southern Canada, between 2013 and 2023. Indicators were defined as ratios of annual maximum LAI or FCOVER between each site and reference forests to control for variability due to climate, species and soils. LAI and FCOVER estimates met Global Climate Observing System uncertainty requirements for over 90% of 94 in-situ validation plots. Indicators gradually (<20% on average) increased in LAI and FCOVER prior to 10 years after planting with LAI continuing to increase (>30% on average) between 10 years and 16 years. Mixedwood sites generally reached reference FCOVER and LAI after 10 years while softwood sites exceeded reference levels by >20% on average. Weak correlations between indicators and climate or site attributes confirm that the reference-based approach effectively normalizes for environmental gradients. Sites with persistently low indicator values were associated with low forest cover due to disturbance or low crown cover due to land degradation, while high-indicator sites represented the upper tail of natural variability rather than discrete anomalies. These findings demonstrate that FCOVER and LAI time series from medium-resolution satellite imagery enable cost-effective, spatially consistent monitoring of afforestation dynamics and identification of poorly regenerating or disturbed sites. • Satellite-based monitoring of 994 afforested sites in a 168,241 km 2 region in Canada. • LAI and vegetation cover indicators from 11 years of Landsat & Sentinel-2 data. • LAI at afforested softwood sites exceeded that of mature forests by 30% on average. • Lowest indicator sites associated with deforestation or low crown cover after 10 yrs.