Abstract. Capturing forest disturbances over time is increasingly important to determine the ecosystem's capacity to recover. Change in disturbance regimes due to climate change increases the frequencies and hinders resilience due to loss of ecological memory and an increase in legacy effects shaping ecosystem recovery. A better understanding of forest disturbances and their role in historical development is needed to develop forest management approaches to promote ecosystem resilience.
Here we use MOD13Q1 data which is highly standardised, ready-to-use time series data for nearly two decades at a global coverage, to detect various known damaged and disturbed areas in forests across Germany to build a chronology of damages and recovery.
Results show that the ability to detect small scale phenomena (such as scattered wind-throw areas), heavily depend on a) the spatial resolution of the data and c) radiometric specifications of the sensor (bands + bandwidth), which also enables us to derive information about event characteristics, and c) the temporal resolution. Difficulties, therefore, still exist in determining the cause of damage for events at supra-pixel resolution (finer than 250m pixel resolution). Ecological phenomena defined by abrupt change e.g. fire or storm are detectable mostly, but limits exist to the differentiation of gradual changing events e.g. insect, drought, or fungi and their interactions. Nevertheless, the analysis captured historic disturbances and allowed the building of plot specific disturbance series over 20 years.