Prevalent trends in realized probability of occurrence of main European forest tree species for 2000–2020
收藏资源简介：
High resolution maps resulting from a trend analysis conducted for the period 2000–2020 on the probability of occurrence maps prepared by Bonannella et al. (2022). For this analysis we selected the realized distribution time series layers at 30m spatial resolution for 6 out of 16 species described in the mentioned publication: Silver fir (<em>Abies alba </em>Mill.) European beech (<em>Fagus sylvatica </em>L.) Norway spruce (<em>Picea abies </em>L.) Black pine (<em>Pinus nigra </em>J. F. Arnold) Scots pine (<em>Pinus sylvestris </em>L.) Common oak (<em>Quercus robur </em>L.) The trend analysis was conducted per pixel on each of these species individually. We fitted simple OLS regression models with the probability of occurrence as the dependent variable and time as the independent variable. After the model fitting, we also calculated the t-test statistics to determine the presence of an increasing (positive) or decreasing (negative) trend or no trend at all. By combining the regression slope coefficient (<em>β</em>) and the <em>p</em>-value from the t-test statistics we assigned each pixel to one of three classes: <em>positive</em>: <em>β</em> > 0.25 AND <em>p</em>-value < 0.05 <em>negative</em>: <em>β</em> < −0.25 AND <em>p</em>-value < 0.05 <em>no trend / stable</em>: −0.25 ≤ <em>β</em> ≥ 0.25 OR <em>p</em>-value > 0.05 We then aggregated the resulting classes at 1km resolution maps to capture the prevalent trend in probability of occurrence over a certain area. Files are named according to the following naming convention, e.g.: veg_abies.alba_slope_30m_0..0cm_epsg3035_v1.0 with the following fields: theme: e.g. <strong>veg</strong>, species code: e.g. <strong>abies.alba</strong>, variable name: e.g. <strong>slope</strong>, resolution in meters e.g. <strong>30m</strong>, reference depths (vertical dimension): e.g. <strong>0..0cm</strong>, coordinate system: e.g. <strong>epsg3035</strong>, data set version: e.g. <strong>v1.0</strong>. For each species here we provide the following layers: veg_abies.alba_<strong>slope</strong>:<strong> </strong>slope coefficient (scaling factor: 10000) veg_abies.alba_<strong>pvalue</strong>:<strong> </strong><em>p</em>-value (scaling factor: 1000) veg_abies.alba_<strong>pos.trends_30m</strong>: pixels classified as <em>positive </em>on the original maps at 30m resolution (boolean layer with range 0–100, only the two extremes values are present) veg_abies.alba_<strong>pos.trends_1km</strong>: proportion of pixels of the <em>positive </em>class over a 1×1 km area (range 0–100) veg_abies.alba_<strong>neg.trends_30m</strong>: pixels classified as <em>negative </em>on the original maps at 30m resolution (boolean layer with range 0–100, only the two extremes values are present) veg_abies.alba_<strong>neg.trends_1km</strong>: proportion of pixels of the <em>negative </em>class over a 1×1 km area (range 0–100) veg_abies.alba_<strong>no.trends_30m</strong>: (pixels classified as <em>no trend / stable </em>on the original maps at 30m resolution (boolean layer with range 0–100, only the two extremes values are present) veg_abies.alba_<strong>no.trends_1km</strong>:<strong> </strong>proportion of pixels of the <em>no trend / stable </em>class over a 1×1 km area (range 0–100) Files are provided as GeoTIFFs and projected in the Coordinate Reference System ETRS89 / LAEA Europe (= EPSG code 3035). Styling files are provided in <em>QML</em> format A publication describing, in detail, all processing steps is currently in review. See at:<br> <br> Bonannella, C., Parente, L., de Bruin, S. and Herold, M. (2023). Multi-decadal trend analysis and forest disturbance assessment of European tree species: concerning signs of a subtle shift, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-3288937/v1]
本数据集为基于Bonannella等人（2022）发布的物种出现概率图集，针对2000—2020年时段开展趋势分析所得到的高分辨率地图。 本次分析选取了该文献中记载的16个树种里的6个，采用其30米空间分辨率的现实分布时间序列图层，分别为：银杉（*Abies alba* Mill.）、欧洲山毛榉（*Fagus sylvatica* L.）、挪威云杉（*Picea abies* L.）、黑松（*Pinus nigra* J. F. Arnold）、欧洲赤松（*Pinus sylvestris* L.）、夏栎（*Quercus robur* L.）。 趋势分析针对每个树种逐像素独立开展。我们构建了以物种出现概率为因变量、时间为自变量的普通最小二乘（Ordinary Least Squares, OLS）回归模型。模型拟合完成后，进一步计算t检验统计量，以判断是否存在上升（正）趋势、下降（负）趋势或无趋势。结合回归斜率系数*β*与t检验得到的*p*值，我们将每个像素划分为以下三类： 正趋势：*β* > 0.25 且 *p*值 < 0.05 负趋势：*β* < −0.25 且 *p*值 < 0.05 无趋势/稳定：−0.25 ≤ *β* ≤ 0.25 或 *p*值 > 0.05 随后我们将分类结果聚合为1公里分辨率的地图，以表征特定区域内物种出现概率的整体趋势。 文件遵循如下命名规范，示例为：`veg_abies.alba_slope_30m_0..0cm_epsg3035_v1.0`，各字段含义如下：主题（如`veg`）、物种代码（如`abies.alba`）、变量名称（如`slope`）、空间分辨率（单位：米，如`30m`）、参考深度（垂直维度，如`0..0cm`）、坐标系（如`epsg3035`）、数据集版本（如`v1.0`）。 针对每个树种，我们提供以下图层： `veg_abies.alba_slope`：回归斜率系数（缩放因子：10000） `veg_abies.alba_pvalue`：*p*值（缩放因子：1000） `veg_abies.alba_pos.trends_30m`：30米分辨率原始图中被归类为正趋势的像素（布尔图层，取值范围0–100，仅包含两个极值） `veg_abies.alba_pos.trends_1km`：1×1公里区域内正趋势类像素占比（取值范围0–100） `veg_abies.alba_neg.trends_30m`：30米分辨率原始图中被归类为负趋势的像素（布尔图层，取值范围0–100，仅包含两个极值） `veg_abies.alba_neg.trends_1km`：1×1公里区域内负趋势类像素占比（取值范围0–100） `veg_abies.alba_no.trends_30m`：30米分辨率原始图中被归类为无趋势/稳定的像素（布尔图层，取值范围0–100，仅包含两个极值） `veg_abies.alba_no.trends_1km`：1×1公里区域内无趋势/稳定类像素占比（取值范围0–100） 数据文件以GeoTIFF格式存储，坐标系采用ETRS89 / LAEA Europe（即EPSG代码3035）。同时提供QML格式的样式文件。 详细阐述所有处理流程的相关论文目前处于审稿阶段，可参阅：Bonannella, C., Parente, L., de Bruin, S. and Herold, M. (2023). 《欧洲树种多年代际趋势分析与森林扰动评估：微妙转变的警示信号》，预印本（版本1），发布于Research Square [https://doi.org/10.21203/rs.3.rs-3288937/v1]
