ForestPaths researchers released a paper titled “Unmixing-based forest recovery indicators for predicting long-term recovery success”, which was published in the Remote Sensing of Environment journal and authored by scientists from the Technical University of Munich in collaboration with Berchtesgaden National Park Research.

While remote sensing is an effective tool for understanding post-disturbance forest recovery, traditional studies often rely on spectral indices, which come with many limitations. They require a long time series after disturbance and do not distinguish between tree regeneration and other post-disturbance vegetation, such as shrubs or grasses. As a result, these indices often underestimate recovery times.

The authors propose spectral unmixing as a more accurate alternative. In the study, they applied spectral unmixing analysis to Landsat and Sentinel-2 fractional cover maps across the eastern Alps (∼130,000 km2) for the period 1990-2021. This allowed them to estimate annual pre- and post-disturbance tree cover and bare ground fractions. Their results showed mean recovery intervals ranging from 5.5 to 13.4 years, depending on the recovery threshold and disturbance severity. Nearly all disturbed areas recovered over the 32-year observation period - 97% by the absolute recovery definition and 95% by the baseline-normalised definition.

The researchers predicted long-term recovery success using data from only three years after a disturbance. Pre-disturbance tree cover, disturbance severity, and post-disturbance bare ground fraction are strong predictors of how much a forest has recovered ten years later. Areas with high pre-disturbance tree cover and low bare ground recovered faster, while the opposite indicated delayed or incomplete recovery. The study highlights that large bare ground fractions are an important early warning sign for areas needing intervention.

Read the full study here.