Monitoring recent forest dynamics in Germany

Abstract

Forests cover about one third of Germany’s land area. They fulfill multiple ecosystem services and play an important role in German economy. Hit by a series of drought periods and heat waves, vast forests areas experienced stress over the past years often resulting in the loss of vitality. While virtually all tree species are affected, the largest impact occurs in coniferous trees. In particular spruce suffered from bark beetle infestations that are usually fatal to the trees. In consequence, tree canopy cover losses occurred at unprecedented rates. Here, we quantify recent forest disturbances and their spatio-temporal dynamics in Germany based on time series of optical Landsat and Sentinel-2 data at monthly resolution. In addition, we derived forest structure (e.g., canopy cover, canopy height, above-ground biomass density) based on GEDI point measurements and Sentinel-1 and Sentinel-2 temporal metrics to demonstrate the effects of recent forest disturbances. Forest structure is expected to have a close relationship to forest biodiversity. In an experiment that simulates a collection of management practices such as stumps, logs, snags, tree removal, crown removal, or habitat trees at small scales, we assessed to which extend these could be detected by means of Sentinel-1 and Sentinel-2 satellite time series. We found that aggregated treatments were generally detectable while distributed treatments did not result in distinct signal changes and remained therefore undetected. Based on the continuous data on canopy cover loss information and forest structure attributes, a comprehensive characterization of pre- and post-disturbance dynamics in Germany is possible. Our results also reveal complex forest structures being resilient to recent climate change conditions. In combination with lessons-learned on spectral indices and spatial statistics detecting the implementation of various aggregated treatments for enhanced structural complexity within a controlled experimental setting, we provide recommendations on specific remotely sensed variables for monitoring. This aids adaptive forest management aimed at boosting structural complexity to foster biodiversity across different scales.