Norway spruce survival during bark beetle outbreak, with tree, stand and climatic variables

The data set comprises tree-, stand-level and environmental attributes of 414 reference and 184 surviving Norway spruce (Picea abies) trees (598 trees in total) that sustained severe bark beetle (Ips typographus) outbreak in the Bohemian Forest region.
The outbreak followed a series of windfalls in the 1980's and is ongoing. The study region spans two spatially disjunct conservation areas, the Bavarian Forest in Germany and Šumava National Park in Czechia. Reference trees refer to living trees sampled from the general population of spruce in the study area. Surviving trees termed "Last Trees Standing", or LTS, were identified using a combination of remote sensing analysis and field surveys. The variables in the data set include tree identification number (treeid), country, tree coordinates, survival status (0 - reference tree, 1 - surviving tree), self-shading ratio, diameter at breast height (dbh) (cm), pre-outbreak stand density (ha), and climatic water balance. Google Earth Pro (GE) 15 cm resolution time series photography available from 2000 to 2019 was used to detect scattered LTS occurrences in the study area. Each image of extant tree was visually and systematically examined to measure projections of the overall shadow length and crown length. Self-shading was calculated as a ratio of crown length to overall tree height derived from measurements of corresponding shadow projections in GE. GE-derived measurements and LTS positions were ground-truthed during field survey. We developed an allometric model to estimate tree diameter at breast height (dbh) from associated GE-derived crown projection area for all LTS. The density of beetle-killed trees within areas of disturbed spruce forest was estimated using spectral and regression analyses. We first classified a Landsat satellite image of the Bohemian Forest to delineate the total areal extent of insect disturbed forest. We then used a regression model to estimate the density of killed trees within the disturbance area. A climatic water balance was calculated as the difference between water supply from precipitation and maximum water losses associated with potential evapotranspiration (PET). PET was estimated using radiation-based method which integrates temperature. We computed global solar radiation using the Area Solar Radiation function in ArcGIS 10.8 and 25 m digital elevation model from Copernicus. Monthly temperature and precipitation values for each year of the 30-year analysis window were derived from two gridded climate datasets: CHELSA (30 arcsec) for the period 1990 to 2013, and E-OBS (0.1°) for 2014 to 2019. The resulting monthly water budget was summed over growing season (May to September) for each year, and averaged over the 30-year analysis period.

本数据集涵盖波西米亚森林区域内遭受严重云杉八齿小蠹（Ips typographus）虫害爆发的598株挪威云杉（Picea abies）的单木水平、林分水平及环境属性数据，其中包含414株参考树与184株存活树。
此次虫害爆发始于20世纪80年代的一系列风倒灾害，目前仍在持续。
研究区域覆盖两个空间上互不相连的保护区：德国巴伐利亚森林国家公园与捷克舒马瓦国家公园。
参考树指从研究区域内云杉总体种群中抽样选取的活立木。被称为“存活保留木（Last Trees Standing，简称LTS）”的存活个体，通过遥感分析与野外调查相结合的方式识别得到。
数据集包含的变量如下：树木编号（treeid）、所属国家、树木坐标、存活状态（0代表参考树，1代表存活树）、自遮阴率、胸径（dbh，单位：厘米）、爆发前林分密度（单位：公顷）以及气候水平衡。
研究人员借助2000年至2019年分辨率达15厘米的谷歌地球专业版（Google Earth Pro, GE）时间序列影像，识别研究区域内零散分布的LTS个体。
研究人员对每株现存树木的影像开展目视与系统检查，以测量总阴影长度与冠幅长度的投影值。自遮阴率通过冠幅长度与总树高的比值计算得出，其中总树高由谷歌地球（GE）中对应阴影投影的测量结果推导得到。
由谷歌地球获取的测量数据与LTS的位置信息，均在野外调查中完成了地面验证。
研究团队构建了异速生长模型，基于所有LTS对应的谷歌地球冠幅投影面积，估算其胸径（dbh）。
借助光谱分析与回归分析，研究人员估算了受干扰云杉林内被甲虫致死的树木密度。具体流程为：首先对波西米亚森林的Landsat卫星影像进行分类，以划定受虫害干扰森林的总分布范围；随后利用回归模型估算该干扰区域内致死树木的密度。
气候水平衡通过降水提供的水分与潜在蒸散量（PET）对应的最大水分损失量之间的差值计算得出。潜在蒸散量（PET）采用整合了温度的辐射法进行估算。
研究团队借助ArcGIS 10.8中的“区域太阳辐射”工具，以及哥白尼计划提供的25米分辨率数字高程模型，计算了全球太阳辐射量。
30年分析周期内各年份的月均气温与降水数据来源于两套网格化气候数据集：1990年至2013年采用CHELSA（30弧秒分辨率），2014年至2019年采用E-OBS（0.1°分辨率）。
最终将各年份生长季（5月至9月）的月水分收支进行求和，并在30年分析周期内取平均值。