Large areas of Europe have been repeatedly affected by severe droughts. Stressed trees suffered from direct drought impacts such as water stress or heat and were also more susceptible to other biotic and abiotic stress agents and calamities. Monitoring such vulnerable forests area-wide is crucial to assess the highly dynamic climate change induced impacts not captured by traditional ground-based monitoring approaches. However, most remote sensing studies dealing with forest condition are either not species-specific, not accounting for morphological and climatic conditions across different regions, not considering natural variations in phenology or not including multiple disturbance agents. Here, we extract species-specific reflectance time series separately for seven natural regions covering Germany for 2016 to 2022. The seasonal evolution of these time series serves as reference for the detection of forest condition anomalies. We calculated a similarity metric – further called forest condition anomaly index () – between each single reflectance observation and the respective measurements within the reference time series, also considering the natural temporal deviations caused by phenology. Temporal aggregation of the allows the generation of spatially comprehensive forest condition anomaly maps. We demonstrate that the shows patterns related to fires, storms and insect infestations and found an overall agreement with state-of-the-art forest disturbance products using a threshold of for forest loss. Consequently, the can be used to detect forest disturbances or linked with vegetation models to assess e.g. forest biomass or carbon flux.

中文翻译：

---

连续树种特定反射率异常指数揭示了 2016 年至 2022 年德国森林状况的恶化


欧洲大片地区多次遭受严重干旱影响。受胁迫的树木遭受直接干旱影响，例如水胁迫或高温，并且也更容易受到其他生物和非生物胁迫因子和灾难的影响。在整个地区监测此类脆弱森林对于评估传统地面监测方法无法捕获的高度动态的气候变化引起的影响至关重要。然而，大多数涉及森林状况的遥感研究要么不针对特定物种，要么不考虑不同地区的形态和气候条件，不考虑物候的自然变化，也不包括多种干扰因素。在这里，我们分别提取了德国七个自然区域2016年至2022年特定物种的反射率时间序列。这些时间序列的季节演变可以为森林状况异常检测提供参考。我们计算了每个单一反射观测值与参考时间序列内的相应测量值之间的相似性度量，进一步称为森林条件异常指数 ()，同时还考虑了物候引起的自然时间偏差。时间聚合允许生成空间综合森林状况异常图。我们证明了这些显示模式与火灾、风暴和虫害相关，并发现与使用森林损失阈值的最先进的森林干扰产品总体一致。因此，它们可用于检测森林干扰或与植被模型联系起来以评估森林生物量或碳通量等。