Abstract
Background Worldwide, forests provide natural resources and ecosystem services. However, forest ecosystems are threatened by increasing forest disturbance dynamics, caused by direct human activities or an altering natural environment. It is decisive to trace the intra- to trans-annual dynamics of these forest ecosystems. National to local forest communities request detailed area-wide maps that delineate forest disturbance dynamics at various spatial scales. Methods We developed a remote sensing based time series analysis (TSA) framework that comprises data access, data management, image pre-processing, and an advanced but flexible TSA. The data basis is a dense time series of multispectral Sentinel-2 images with a spatial resolution of 10 metres. We use a dynamic Savitzky-Golay-filtering approach to reconstruct robust but sensitive phenology courses. Deviations from the latter are further used to derive spatiotemporal information on forest disturbances. In a first case study, we apply the TSA to map forest disturbances directly or indirectly linked to recurring bark beetle infestation in Northern Austria. Finally, we use zonal statistics on different spatial scales to provide aggregated information on the extent of forest disturbances between 2018 and 2019.Results and Conclusion The outcomes are a) individual phenology models and deduced phenology metrics for each 10 metres by 10 metres forest pixel in Austria and b) forest disturbance maps useful to investigate the occurrence, development and extent of bark beetle infestation. The phenology modelling results provide area-wide consistent data, also useful for downstream analyses (e.g. forest type classification). Results of the forest disturbance detection demonstrate that the TSA is capable to systematically delineate disturbed forest areas. Information derived from such a forest monitoring tool is highly relevant for various stakeholders in the forestry sector, either for forest management purposes or for decision-making processes on different levels.
Approximation of the disturbance dynamics by Extended State Observer Using an Artificial Delay
