ECOCLIMAP-SG+: an agreement-based high-resolution land use land cover dataset for meteorological modelling
收藏资源简介：
ECOCLIMAP-SG+ is a new 60 m land use land cover dataset, which covers a continental domain, and represents the 33 labels of the original ECOCLIMAP-SG dataset. ECOCLIMAP-SG is used in HARMONIE-AROME, the numerical weather prediction model used operationally by Met Éireann and other national meteorological services. ECOCLIMAP-SG+ was created using an agreement-based method to combine information from many maps to overcome variations in semantic and geographical coverage, resolutions, formats, accuracies, and representative periods. In addition to ECOCLIMAP-SG+, the process generates an agreement score map, which estimates the uncertainty of the land cover labels in ECOCLIMAP-SG+ at each location in the domain.Using a set of primary labels, ECOCLIMAP-SG+ outperforms ECOCLIMAP-SG regarding the F1-score against LUCAS 2022 over Europe and the Irish national land cover 2018 dataset. Similarly, it outperforms the Icelandic version of ECOCLIMAP-SG for most of the represented secondary labels. ECOCLIMAP-SG+ can be further improved by including new maps once they become available. Code: The code used to create ECOSG+ is available on Gitlab. The notebook example/Create\_download\_mstar.ipynb and example/Assemble\_maps.ipynb can be ran in Google Colab but one must first sign up for a Google Earth Engine account to run the code. The assembly code requires ECOCLIMAP-SG which is available through CNRM Contents: Each zip file contains 100 tiles covering the EURAT domain (longitudes: -32 to 42 degrees, latitudes: 20 to 72 degrees). Each tif file has a resolution of 60m. quality_score_map.zip: The agreement quality score. best-guess_map.zip : The so-called best-guest map which represents the label with the highest quality score for each pixel. ecosg_plus.zip: ECOCLIMAP-SG+ map over the Eurat domain divided in 100 .tif files degrees). This map is a composition of the best-guess map where the quality score is greater than 52.5% and ECOCLIMAP-SG everywhere else. Labels: 1. Sea and oceans 2. Lakes 3. Rivers 4. Bare land 5. Rocks 6. Permanent snow and ice 7. Boreal broadleaf deciduous 8. Temperate broadleaf deciduous 9. Tropical broadleaf deciduous 10 .Temperate broadleaf evergreen 11. Tropical broadleaf evergreen 12. Boreal needleleaf evergreen 13. Temperate needleleaf evergreen 14. Boreal needleleaf deciduous 15. Shrubs 16. Boreal grassland 17 .Temperate grassland 18. Tropical grassland 19. Winter C3 crops 20. Summer C3 crops 21.C4 crops 22. Flooded trees 23. Flooded grassland 24. LCZ1: compact high-rise 25. LCZ2: compact mid rise 26. LCZ3: compact low-rise 27. LCZ4: open high-rise 28. LCZ5: open midrise 29. LCZ6: open low-rise 30 .LCZ7: lightweight low-rise 31. LCZ8: large low-rise 32. LCZ9:sparsely built 33. LCZ10: heavy industry
ECOCLIMAP-SG+是一款全新的60米分辨率土地利用与土地覆盖数据集，覆盖全域大陆范围，完整涵盖原始ECOCLIMAP-SG数据集的33类地表覆盖标签。原始ECOCLIMAP-SG数据集已被应用于HARMONIE-AROME数值天气预报模式——该模式由爱尔兰气象局（Met Éireann）及其他多国气象业务部门正式投入业务运行。 ECOCLIMAP-SG+采用基于一致性融合的方法构建：通过整合多源地图数据，以解决不同数据集在语义与地理覆盖范围、空间分辨率、数据格式、精度以及代表时段上存在的差异。除生成ECOCLIMAP-SG+数据集外，该构建流程还同步产出一张一致性评分图，用于估算研究区域内每个像素点对应的ECOCLIMAP-SG+地表覆盖标签的不确定性。 基于一套核心标签体系，相较于原始ECOCLIMAP-SG数据集，ECOCLIMAP-SG+在欧洲区域针对LUCAS 2022数据集以及爱尔兰2018年全国土地覆盖数据集的F1分数上均表现更优。同理，针对绝大多数次级地表覆盖标签，ECOCLIMAP-SG+的表现也优于ECOCLIMAP-SG的冰岛版本。后续可通过纳入新的公开地图数据，进一步优化ECOCLIMAP-SG+数据集的性能。 **Code:** ECOSG+的构建代码已托管至Gitlab平台。示例脚本`example/Create_download_mstar.ipynb`与`example/Assemble_maps.ipynb`可在谷歌Colab（Google Colab）中运行，但需先注册谷歌地球引擎（Google Earth Engine）账号后方可执行代码。该拼接代码依赖ECOCLIMAP-SG数据集，此数据集可通过法国国家气象研究中心（CNRM）获取。 **Contents:** 每个压缩包均包含100个影像瓦片，覆盖EURAT研究区域（经度范围：-32°至42°，纬度范围：20°至72°），单个TIFF文件的空间分辨率为60米。 quality_score_map.zip：一致性质量评分数据集。 best-guess_map.zip：即所谓最优猜测图，为研究区域内每个像素匹配质量评分最高的地表覆盖标签。 ecosg_plus.zip：覆盖EURAT研究区域的ECOCLIMAP-SG+数据集，被切分为100个TIFF文件。该数据集由两部分拼接而成：质量评分高于52.5%的区域采用最优猜测图的标签，其余区域则沿用原始ECOCLIMAP-SG数据集的标签。 **Labels:** 1. 海洋 2. 湖泊 3. 河流 4. 裸地 5. 岩石 6. 永久积雪与冰川 7. 寒温带落叶阔叶林 8. 温带落叶阔叶林 9. 热带落叶阔叶林 10. 温带常绿阔叶林 11. 热带常绿阔叶林 12. 寒温带常绿针叶林 13. 温带常绿针叶林 14. 寒温带落叶针叶林 15. 灌丛 16. 寒温带草原 17. 温带草原 18. 热带草原 19. 冬季C3作物 20. 夏季C3作物 21. C4作物 22. 洪泛森林 23. 洪泛草原 24. LCZ1：密集高层建筑区 25. LCZ2：密集中层建筑区 26. LCZ3：密集低层建筑区 27. LCZ4：开阔高层建筑区 28. LCZ5：开阔中层建筑区 29. LCZ6：开阔低层建筑区 30. LCZ7：轻型低层建筑区 31. LCZ8：大型低层建筑区 32. LCZ9：零散建筑区 33. LCZ10：重工业区
