Occurrence of forest ecosystem services

Dataset title: Occurrence of forest ecosystem services
Theme: Environment and Conservation
Description: Wood

Basic information: the data shows the 20% of forests supplying the largest amount of timber on a per unit area basis. Timber provision is measured in terms of growing stock volume (m3/ha). Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Water supply
Basic information: the data shows the 20% of forests supplying the largest amount of freshwater on a per unit area basis. Water supply is measured as mm/year. Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Erosion control
Basic information: the data shows the 20% of forests providing the most powerful action of erosion control on a per unit area basis. Erosion control is measured as avoided soil erosion (tons/ha year). Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Pollination
Basic information: the data shows the 20% of forests showing the strongest pollination potential on a per unit area basis. Pollination potential is a qualitative indicator measuring the capacity of ecosystems, including forest edges, to support pollinator insects. Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Habitat provision
Basic information: the data shows the 20% of forests with the strongest relative potential as a habitat for bird species. Habitat provision is expressed as the ratio (%) between local bird species richness and the average richness in the regional context (250-km radius). Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Soil formation
Basic information: the data shows the 20% of forest soils storing the largest amount of organic carbon on a per unit area basis. Soil organic carbon is intended as the content of organic carbon in both the topsoil (0-30 cm) and the subsoil (30-100 cm), and is measured in tons/ha. Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Climate regulation
Basic information: the data shows the 20% of forests storing the largest amount of above- and below-ground carbon on a per unit area basis. The data, which is expressed in tons/ha, accounts for forest type (coniferous or broadleaved) and ecological zone. Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Recreation
Basic information: the data shows the 20% of forests guaranteeing the highest recreation potential on a per unit area basis. The recreation potential is a qualitative indicator accounting for recreational quality and accessibility. Pixels with a value of 1 identify the top suppliers, whereas pixels with a value of 0 identify other forested land parcels (i.e. bottom 80%).

Data processing steps: processing
Description: Wood

Data processing step: the data was extracted from the relevant dataset produced in the GlobBiomass project (Santoro et al., 2018). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Water supply
Data processing step: freshwater supply was estimated as the amount of water running off each land parcel in the landscape owing to the combined effect of precipitation and evapotranspiration, using the Annual Water Yield model of InVEST 3.3 (Integrated Valuation of Ecosystem Services and Tradeoffs). Inouts for the modeling were obtained as explained below. Data about average annual precipitation at each location were extracted from the WorldClim global climate database (http://www.worldclim.org). The average annual reference evapotranspiration was obtained from the Global Reference Evapotranspiration Version 2 dataset (Trabucco and Zomer, 2019). The root restricting layer depth was obtained from the JRC European Soil Data Centre’s (ESDAC) European Soil Database (Panagos et al., 2012; Hiederer, 2013). The plant available water content was derived from the relevant map (Available Water Capacity) in the “Topsoil physical properties for Europe” database by ESDAC (Panagos et al., 2012; Ballabio et al., 2016). Land use/land cover information was directly imported from the Corine land cover, while watersheds were delineated from the European Digital Elevation Model (EU-DEM) using GIS-based hydrological tools. The maximum root depth per vegetated land use class was extracted from the Forest and Agriculture Organization (FAO) guidelines (Allen et al., 1998) and Schenk and Jackson (2002). The vapotranspiration coefficient (Kc) was extracted from Allen et al. (1998) and Nistor et al. (2017) for vegetated areas, and from InVEST guidelines for urban areas and water bodies. Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Erosion control
Data processing step: the data was extracted from the relevant dataset produced by the Joint Research Centre (JRC) (Maes et al., 2015; Guerra et al., 2016). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Pollination
Data processing step: the data was extracted from the Relative Pollination Potential (RPP) map produced by the Joint Research Centre (JRC) (Zulian et al., 2013). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Habitat provision
Data processing step: the data was extracted from the JRC’s Habitat Quality dataset (Vallecillo et al., 2016). A resampling was performed to bring the JRC dataset from a 10-km resolution to a 1-km one. Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Soil formation
Data processing step: the data was extracted from the amended Harmonised World Soil Database (Hiederer and Köchy, 2012). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Climate regulation
Data processing step: the data was extracted from the Pan-European dataset of forest above- and below-ground carbon produced by the JRC (de Rigo et al., 2013). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Recreation
Data processing step: the data was obtained through a reclassification of the Recreation Opportunity Spectrum (ROS) map produced by the JRC at a 100-m resolution (Paracchini et al., 2014). The reclassification was based on the assumption that the supply of recreation opportunities is proportional to the product of the two ROS variables (i.e. quality and accessibility). The reclassified map was made continuous by running a neighborhood operation assigning each pixel the mean of pixels within a radius of 1000 m. A nearest neighbor resampling was finally performed to bring the resolution to 1 km. Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Data processing steps processing
Description Wood
Data processing step: the data was extracted from the relevant dataset produced in the GlobBiomass project (Santoro et al., 2018). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Water supply
Data processing step: freshwater supply was estimated as the amount of water running off each land parcel in the landscape owing to the combined effect of precipitation and evapotranspiration, using the Annual Water Yield model of InVEST 3.3 (Integrated Valuation of Ecosystem Services and Tradeoffs). Inouts for the modeling were obtained as explained below. Data about average annual precipitation at each location were extracted from the WorldClim global climate database (http://www.worldclim.org). The average annual reference evapotranspiration was obtained from the Global Reference Evapotranspiration Version 2 dataset (Trabucco and Zomer, 2019). The root restricting layer depth was obtained from the JRC European Soil Data Centre’s (ESDAC) European Soil Database (Panagos et al., 2012; Hiederer, 2013). The plant available water content was derived from the relevant map (Available Water Capacity) in the “Topsoil physical properties for Europe” database by ESDAC (Panagos et al., 2012; Ballabio et al., 2016). Land use/land cover information was directly imported from the Corine land cover, while watersheds were delineated from the European Digital Elevation Model (EU-DEM) using GIS-based hydrological tools. The maximum root depth per vegetated land use class was extracted from the Forest and Agriculture Organization (FAO) guidelines (Allen et al., 1998) and Schenk and Jackson (2002). The vapotranspiration coefficient (Kc) was extracted from Allen et al. (1998) and Nistor et al. (2017) for vegetated areas, and from InVEST guidelines for urban areas and water bodies. Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Erosion control
Data processing step: the data was extracted from the relevant dataset produced by the Joint Research Centre (JRC) (Maes et al., 2015; Guerra et al., 2016). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Pollination
Data processing step: the data was extracted from the Relative Pollination Potential (RPP) map produced by the Joint Research Centre (JRC) (Zulian et al., 2013). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Habitat provision
Data processing step: the data was extracted from the JRC’s Habitat Quality dataset (Vallecillo et al., 2016). A resampling was performed to bring the JRC dataset from a 10-km resolution to a 1-km one. Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Soil formation
Data processing step: the data was extracted from the amended Harmonised World Soil Database (Hiederer and Köchy, 2012). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Climate regulation
Data processing step: the data was extracted from the Pan-European dataset of forest above- and below-ground carbon produced by the JRC (de Rigo et al., 2013). Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Recreation
Data processing step: the data was obtained through a reclassification of the Recreation Opportunity Spectrum (ROS) map produced by the JRC at a 100-m resolution (Paracchini et al., 2014). The reclassification was based on the assumption that the supply of recreation opportunities is proportional to the product of the two ROS variables (i.e. quality and accessibility). The reclassified map was made continuous by running a neighborhood operation assigning each pixel the mean of pixels within a radius of 1000 m. A nearest neighbor resampling was finally performed to bring the resolution to 1 km. Top 20% supplying pixels were given a value of 1, whereas all other pixels were given a value of 0.

Additional information (URL): innoforest.eu

数据集标题：森林生态系统服务供给数据集
主题：环境与保护
核心描述：木材供给
基本信息：该数据集展示了单位面积木材供给量最高的20%森林区域。木材供给量以立木蓄积量（m³/公顷）为计量单位。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

供水服务
基本信息：该数据集展示了单位面积淡水供给量最高的20%森林区域。淡水供给量以毫米/年（mm/年）为计量单位。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

侵蚀控制
基本信息：该数据集展示了单位面积土壤侵蚀控制效能最高的20%森林区域。土壤侵蚀控制效能以避免土壤侵蚀量（吨/公顷·年）为计量单位。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

授粉服务
基本信息：该数据集展示了单位面积传粉潜力最高的20%森林区域。传粉潜力为定性指标，用于衡量包括林缘在内的生态系统支撑传粉昆虫的能力。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

栖息地供给
基本信息：该数据集展示了作为鸟类物种栖息地相对潜力最高的20%森林区域。栖息地供给能力以半径250公里的区域内本地鸟类物种丰富度与区域平均物种丰富度的比值（%）表示。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

土壤形成
基本信息：该数据集展示了单位面积土壤有机碳储量最高的20%森林土壤。土壤有机碳指表层土壤（0-30cm）与亚表层土壤（30-100cm）中的有机碳含量，计量单位为吨/公顷。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

气候调节
基本信息：该数据集展示了单位面积地上与地下碳储量最高的20%森林区域。该数据以吨/公顷为计量单位，纳入了森林类型（针叶林或阔叶林）与生态区的影响。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

游憩服务
基本信息：该数据集展示了单位面积游憩潜力最高的20%森林区域。游憩潜力为定性指标，综合考量了游憩质量与可达性。像素值为1的区域为顶级供给区，像素值为0的区域为其余森林地块（即后80%森林）。

数据处理流程
木材供给
数据处理步骤：数据提取自GlobBiomass项目（Santoro等，2018）生成的相关数据集。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

供水服务
数据处理步骤：采用生态系统服务与权衡综合评估模型（Integrated Valuation of Ecosystem Services and Tradeoffs，简称InVEST）3.3的年度产水量模型，基于降水与蒸散发的共同作用，估算景观中每个地块的径流量，以此作为淡水供给量。建模所需输入数据如下：
1. 各点位年平均降水量数据提取自WorldClim全球气候数据库（http://www.worldclim.org）；
2. 年平均参考蒸散发量提取自《全球参考蒸散发版本2》数据集（Trabucco与Zomer，2019）；
3. 根系限制层深度提取自联合研究中心（Joint Research Centre, JRC）欧洲土壤数据中心（ESDAC）的欧洲土壤数据库（Panagos等，2012；Hiederer，2013）；
4. 植物有效含水量源自ESDAC的“欧洲表层土壤物理性质”数据库中的相关地图（有效水容量）（Panagos等，2012；Ballabio等，2016）；
5. 土地利用/土地覆盖信息直接导入自Corine土地覆盖数据，流域边界则基于欧洲数字高程模型（EU-DEM），通过GIS水文工具提取；
6. 各植被土地利用类别的最大根深提取自联合国粮食及农业组织（Food and Agriculture Organization, FAO）指南（Allen等，1998）与Schenk和Jackson（2002）；
7. 植被区的蒸散发系数（Kc）提取自Allen等（1998）与Nistor等（2017），城镇与水体区域的蒸散发系数则参考InVEST指南。
最终将供给量排名前20%的像素赋值为1，其余像素赋值为0。

侵蚀控制
数据处理步骤：数据提取自联合研究中心（JRC）生成的相关数据集（Maes等，2015；Guerra等，2016）。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

授粉服务
数据处理步骤：数据提取自联合研究中心（JRC）生成的相对传粉潜力（Relative Pollination Potential, RPP）地图（Zulian等，2013）。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

栖息地供给
数据处理步骤：数据提取自JRC的栖息地质量数据集（Vallecillo等，2016）。通过重采样将JRC数据集的分辨率从10公里提升至1公里。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

土壤形成
数据处理步骤：数据提取自修正后的 harmonised World Soil Database（Hiederer与Köchy，2012）。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

气候调节
数据处理步骤：数据提取自JRC生成的泛欧洲森林地上与地下碳储量数据集（de Rigo等，2013）。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

游憩服务
数据处理步骤：数据通过对JRC生成的100米分辨率游憩机会谱（Recreation Opportunity Spectrum, ROS）地图进行重分类得到（Paracchini等，2014）。重分类基于以下假设：游憩机会供给量与ROS的两个变量（即质量与可达性）的乘积成正比。通过邻域运算对重分类后的地图进行连续化处理：为每个像素赋值为其1000米半径范围内所有像素的平均值。最终通过最近邻重采样将分辨率调整至1公里。将供给量排名前20%的像素赋值为1，其余像素赋值为0。

附加信息（网址）：innoforest.eu