Global Wind Atlas v3 (orography and bathymetry)

This data combines the GEBCO bathymetric dataset with the SRTMv3 and Viewfinder 3 arc-second orography data. The data is interpolated to 250m, and all land locations are given positive values, while water surfaces are given negative values.<br><b>Orographic height over land</b><br>Void-filled SRTM data was used between 60°N and 60°S, with a few exceptions where the void filling process introduced artifacts that were not present in the Viewfinder Panoramas Digital Elevation Model. This was particularly apparent in the Pacific islands.<br><br>North of 60°N the Viewfinder Panoramas DEM was used, allowing us to include model results for large parts of Scandinavia, Russia, and Canada. For these areas, the main sources to the Viewfinder Panoramas DEM are height contours from Russian military maps and Canadian Digital Elevation Data. The Viewfinder Panoramas DEM had several void regions which were filled using a cubic interpolation in the y-dimension of the raster. This was chosen after investigating many of the void regions and finding that they tended to run east-west.<br>Both DEM datasets were provided as 1° by 1° tiles in the WGS 1984 coordinate system (EPSG: 4326). To support the UTM projection used in the modeling, the data was re-projected to a 150m grid spacing using cubic interpolation. The interpolation was done using the Geospatial Data Abstraction Library (GDAL) tool gdalwarp. The 150m resolution was selected as it corresponds to the effective resolution of the SRTM data.<br>The data was then re-projected again in the WAsP model to the 250m using the approach of the European Wind Atlas.<br><br><b>Bathymetry data (Oceans)</b><b><br></b>Negative elevation data for seabed bathymetry was obtained from GEBCO’s gridded bathymetric data set, the GEBCO_2020 grid, which is a global terrain model for ocean and land at 15 arc-second intervals. The GEBCO_2020 grid was prepared for use in the GWA by the World Bank Group.<br>Data set attribution: GEBCO Compilation Group (2020) GEBCO 2020 Grid (doi:10.5285/a29c5465-b138-234d-e053-6c86abc040b9).<br><b>Data combination</b><b><br></b>The datasets were combined by first masking each using the fine resolution of the GSHHG coastline dataset, to ensure that all land values were positive and all water values were negative, with no data have a value of 0.

本数据集融合了GEBCO水深数据集（GEBCO bathymetric dataset）与SRTMv3及Viewfinder 3弧秒地形高程数据。所有数据均被重采样至250米分辨率；其中陆地区域赋值为正值，水体表面赋值为负值。

**陆地地形高程**
在南北纬60°之间的区域采用经空洞填充的SRTM数据，但存在少量例外情况：空洞填充过程引入了Viewfinder全景数字高程模型（Viewfinder Panoramas Digital Elevation Model，简称DEM）中未出现的伪影，此类问题在太平洋岛屿区域尤为显著。

在北纬60°以北区域，我们采用了Viewfinder全景DEM，借此可覆盖斯堪的纳维亚半岛、俄罗斯及加拿大的大部分区域。该DEM的主要数据源为俄罗斯军用地图高程等高线与加拿大数字高程数据。Viewfinder全景DEM存在多处空洞区域，我们通过对栅格的y轴方向进行三次插值完成空洞填充——这一方案是在对大量空洞区域进行调研后确定的，因观测到此类空洞多呈东西走向。

两类DEM数据集均以1°×1°瓦片的形式提供，坐标系采用WGS 1984（EPSG: 4326）。为适配建模中使用的通用横轴墨卡托（UTM）投影，我们通过三次插值将数据重投影至150米的网格间距，插值操作借助地理空间数据抽象库（Geospatial Data Abstraction Library, GDAL）的gdalwarp工具完成。选择150米分辨率是因其与SRTM数据的有效分辨率相匹配。

随后，借助欧洲风能图集（European Wind Atlas）的方法，我们在WAsP模型中将数据再次重投影至250米分辨率。

**海洋水深数据**
海底水深的负高程数据源自GEBCO的格网化水深数据集——GEBCO_2020格网，该数据集是一套以15弧秒为间隔的全球海陆地形模型。GEBCO_2020格网由世界银行集团为全球风能图集（GWA）的应用进行预处理。

数据集引用信息：GEBCO汇编工作组（2020）GEBCO 2020格网（doi:10.5285/a29c5465-b138-234d-e053-6c86abc040b9）。

**数据融合流程**
数据集融合时，首先借助高分辨率的GSHHG海岸线数据集对两类数据分别进行掩膜处理，以确保所有陆地区域的赋值为正、水体区域赋值为负，无数据区域的赋值为0。