Reconstructing forest dynamics in the European Alps through a high-dimensional analysis based on Landsat time series

Natural disturbances are increasingly threatening forest ecosystems due to climate change globally. In
Europe, disturbance regimes have intensified over the last decades, leading to increased size, frequency
and severity of disturbance events. Satellite remote sensing data acquired over the past decades are crucial
for assessing changes in disturbance regimes as they provide wall-to-wall spatial information from the
landscape to the global scale. In particular, Landsat imagery has been continuously acquired since 1984,
and it offers an unprecedented opportunity for mapping land cover changes thanks to its spatial and
spectral consistency. Following the opening of the USGS Landsat archive, dense time series have been
exploited through automated algorithms for targeting forest dynamics. Currently, the most widely used
algorithms aim to detect abrupt and gradual changes by performing a temporal segmentation of Landsat
time series at the pixel level. The sensitivity of automated algorithms has been enhanced by including
multiple spectral and spatial information in the time series though their combined usage is still limited.
Here, we present an automated algorithm for detecting forest dynamics named High-dimensional detection
of Land Dynamics (HILANDYN), which exploits the temporal, spatial and spectral dimensions of inter-annual
Landsat time series. We tested HILANDYN to map forest disturbances that occurred during the last four
decades in the European Alps. HILANDYN builds upon a statistical procedure for detecting changepoints in
high-dimensional time series through a bottom-up segmentation procedure. Our results showed that the
algorithm is sensitive toward a wide range of disturbance severities and can detect stand-replacing events,
e.g. windthrows and wildfires, and non-stand-replacing ones, e.g. insect outbreaks and drought-induced
dieback. Moreover, we were able to map disturbances occurring in consecutive years, such as windthrows
followed by salvage logging. We obtained the best results in terms of accuracy metrics using a combination
of original bands and indices that included the heterogeneous spectral information provided by the
multispectral sensors of the Landsat missions. In particular, we achieved an F1 score equal to 83% for the
disturbed class, corresponding to a user’s accuracy of 84,3% and a producer’s accuracy of 82%. Accurate
disturbance maps of the European Alps will enable a thorough analysis of the shifts in the disturbance
regimes over the last four decades, alongside the assessment of forest recovery patterns under different
management practices and environmental conditions.