Winter reddening of young Douglas-fir (Pseudotsuga menziesii Mirb. Franco), triggered by large thermal fluctuations in late winter, is a critical problem for European forestry. A literature review identified certain climatic conditions that are characteristic of 'reddening' years, including warm daily temperatures, high daily temperature amplitude, low relative humidity, moderate wind speeds, as well as the occurrence of freeze-thaw cycles with cold night temperatures. By describing the triggering environmental and stand factors, we propose three hypotheses for the physiological processes leading to winter reddening, namely (i) hydraulic failure due to winter drought stress, (ii) photo-oxidative stress in shade-acclimated trees, and (iii) early cold deacclimation during warm periods. i) Low soil temperature, by reducing root water uptake, combined with anticyclonic conditions, by increasing water losses, can induce hydraulic failure in the xylem. Hydraulic failure may be further accelerated by night frosts. ii) Winter reddening can occur when low temperature and high irradiance coincide, disrupting photostasis. Overwhelming of winter photo-protection may lead to photodamage and subsequent reddening. iii) Warm periods, by inducing cold deacclimation, make trees susceptible to frost damage. Finally, the three processes may interact under atypical anticyclonic conditions in late winter (e.g. cold or dry soils, warm days, high irradiance and/or freezing nights). Indeed, trees under water stress would develop a higher sensitivity to freezing night and photooxidative stress. We therefore proposed mitigation actions to avoid exposing trees to stressful conditions based on e.g. stand characteristics, understorey vegetation and planting.

Introduction: Soil drought during summer in Central Europe has become more frequent and severe over the last decades. European forests are suffering increasing damage, particularly Norway spruce. Douglas-fir (Pseudotsuga menziesii (Mirbel) Franco), a non-native tree species, is considered as a promising alternative to build drought-resilient forests. The main goal of this study was to investigate the intraannual radial stem growth and sap flow performance of Douglas-fir along a precipitation gradient across Germany under severe drought. Material and methods: Sap flow and stem radial changes of up to ten trees each at four sites with different precipitation regimes were measured in combination with volumetric soil water content during the growing season of 2022. Measurements of stem radial changes were used to calculate the trees' stem water deficit, a proxy for tree water status and drought stress. Results: The severe summer drought of 2022 led to an early growth cessation and a significant reduction in daily sap flow at all four sites monitored. We could identify a site-specific threshold in soil water availability ranging between 21.7 and 29.6% of relative extractable water (REW) under which stem water reserves cannot be replenished and thereby inhibiting radial growth. We could also demonstrate that at this threshold, sap flow is heavily reduced to between 43.5 and 53.3%, and for a REW below 50%, sap flow linearly decreases by 1.1-2.0% per 1% reduction in REW. This reduction tends to follow the humidity gradient, being more pronounced at the most oceanic characterized site and suggesting an adaptation to site conditions. Even though Douglas-fir is considered to be more drought stress resistant than Norway spruce, growth and sap flow are greatly reduced by severe summer drought, which became more frequent in recent years and their frequency and intensity is likely to increase. Conclusions: Our results suggest that timber production of Douglas-fir in Central Europe will decline considerably under projected climate change, and thus pointing to site specific growth constraints for a so far promising non-native tree species in Europe.

冰湖是气候变化的重要指示器，在全球变暖的背景下，冰湖的持续监测对区域水资源管理以及高山地区防灾减灾具有重要意义。受冰川消融和气候影响，冰湖边界会发生季节性和年际变化。已有冰湖制图的研究，往往要首先确定遥感影像上每个冰湖的位置，然后获取其精细边界。近年来，不同区域的冰湖编目数据日益增加，提供了大量的冰湖历史边界。对这些位置已知的冰湖进行监测只需提取其当前边界，而历史边界可以作为冰湖迭代边界的起点，从而加速冰湖编目数据的更新。本研究以喜马拉雅山脉成像条件较好的488个冰湖和受积雪、冰冻、云、山体阴影影响的80个冰湖作为研究对象，将前者按照面积大小分成3类，基于1990年冰湖编目数据提供的历史边界信息，对比了人工阈值法、OTSU阈值法、U-NET、双峰迭代法、OTSU迭代法和C-V迭代法在2014年后Landsat-8 OLI影像上的冰湖提取结果。结果表明：OTSU迭代法、C-V迭代法能有效利用冰湖缓冲区内的统计信息，取得的F1分数高达88.89%和89.30%，显著优于人工阈值法、OTSU阈值法和双峰迭代法，也能较完整地提取冰冻状态和云覆盖的冰湖；对于积雪覆盖下的冰湖，C-V迭代法的提取精...

冰湖是气候变化的重要指示器，在全球变暖的背景下，冰湖的持续监测对区域水资源管理以及高山地区防灾减灾具有重要意义。受冰川消融和气候影响，冰湖边界会发生季节性和年际变化。已有冰湖制图的研究，往往要首先确定遥感影像上每个冰湖的位置，然后获取其精细边界。近年来，不同区域的冰湖编目数据日益增加，提供了大量的冰湖历史边界。对这些位置已知的冰湖进行监测只需提取其当前边界，而历史边界可以作为冰湖迭代边界的起点，从而加速冰湖编目数据的更新。本研究以喜马拉雅山脉成像条件较好的488个冰湖和受积雪、冰冻、云、山体阴影影响的80个冰湖作为研究对象，将前者按照面积大小分成3类，基于1990年冰湖编目数据提供的历史边界信息，对比了人工阈值法、OTSU阈值法、U-NET、双峰迭代法、OTSU迭代法和C-V迭代法在2014年后Landsat-8 OLI影像上的冰湖提取结果。结果表明：OTSU迭代法、C-V迭代法能有效利用冰湖缓冲区内的统计信息，取得的F1分数高达88.89%和89.30%，显著优于人工阈值法、OTSU阈值法和双峰迭代法，也能较完整地提取冰冻状态和云覆盖的冰湖；对于积雪覆盖下的冰湖，C-V迭代法的提取精...