本数据集基于MCD19A2、MERRA-2、ERA5_land数据集，通过数据预处理、数据上采样、时空匹配将气溶胶光学厚度（AOD）、波长指数（AE）、气象参数等多种数据提取，将AERONET地面站点观测值作为目标真值，利用机器学习算法LightGBM训练模型，反演2000年至2022年“一带一路”沙漠地区的逐日1公里分辨率陆地沙尘气溶胶光学厚度。本次沙尘气溶胶光学厚度数据集为“一带一路”地区的沙尘研究提供基础数据。\n
1. 数据集命名：Dust_Aerosol_Optica_Depth\n
2. 属性信息\n
Lon：数据集的经度(°)\n
Lat:：数据集的纬度 (°)\n
Dust_AOD：沙尘气溶胶光学厚度", "ds_source": "
1. MCD19A2数据\n
从美国MODIS官方网站(https://modis.gsfc.nasa.gov//) 下载。\n
2. MERRA-2 tavg1_2d_aer_Nx数据\n
从美国宇航局戈达德地球科学数据和信息服务中心(https://disc.gsfc.nasa.gov/) 下载。（该数据需要下载全球范围）\n
3. ERA5-Land数据\n
从欧洲中期天气预报中心(https://cds.climate.copernicus.eu/) 下载。（该数据需要下载东经0°-东经150°，北纬81°-南纬11°的范围）", "ds_process_way": "
1. 遥感数据预处理：MERRA-2(0.5°×0.625°)和ERA5-Land(0.1°×0.1)分别通过双线性插值采样至1公里分辨率，其中ERA5-Land在此之前还需通过最近邻插值填补缺失值，按照最近点像素的方式时空匹配，提取所需数据。将研究范围内的AERONET地面站点观测数据，筛选出波长指数(AE)<1的气溶胶光学厚度(AOD)作为沙尘AOD的目标真值。\n
2. 反演方法：采用最终采用LightGBM模型分别训练两个模型：(1)完整模型，即各项预测因子均未缺失作为输入训练的模型（MCD19A2数据存在大量缺失，针对非缺失部分）；(2)非完整模型，即MCD19A2存在数据缺失的部分，仅采用由MERRA-2、ERA5-Land数据提取的预测因子为输入训练的模型。两个模型的输出需要与经降尺度后的MERRA-2的沙尘AOD数据取均值处理，才作为最终反演结果。\n
参考文献：Ke G, Meng Q, Finley T, et al. Lightgbm: A highly efficient gradient boosting decision tree[J]. Advances in neural information processing systems, 2017, 30.", "ds_quality": "
取AERONET站点经匹配后未用于训练的数据作验证。平均绝对误差（MAE）为0.0515；均方根误差（RMSE）为0.0805；平均偏差（MBE）为-0.0252；相关系数（R）为0.8559；落在期望误差（EE）为81.74%。", "ds_acq_start_time": "2000-01-01 00:00:00", "ds_acq_end_time": "2022-12-31 00:00:00", "ds_acq_place": "“一带一路”沿线沙漠地区", "ds_acq_lon_east": 117.0, "ds_acq_lat_south": 21.0, "ds_acq_lon_west": 62.0, "ds_acq_lat_north": 49.0, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 3030931197773, "ds_files_count": 175321, "ds_format": "tif", "ds_space_res": "1km", "ds_time_res": "日", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "f48b3581-bdc3-43ad-8a96-fc9115628b4d.jpg", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": null, "organization_id": "9de89acc-5714-4927-aba3-ac88067dff8a", "ds_serv_man": "敏玉芳", "ds_serv_phone": "0931-4967596", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [ "170.4510" ], "quality_level": 3, "publish_time": "2025-04-14 18:22:45", "last_updated": "2025-12-08 18:57:23", "protected": false, "protected_to": null, "lang": "zh", "cstr": "11738.11.NCDC.NIEER.DB6814.2025", "i18n": { "en": { "title": "Daily dust aerosol optical thickness dataset along the \"the Belt and Road\" (2000-2022)", "ds_format": "tif", "ds_source": "
1. MCD19A2 data\n
From MODIS official website (https://modis.gsfc.nasa.gov//) to download.\n
2. MERRA-2 tavg1_2d_aer_Nx data\n
From NASA's Goddard earth science data and information service center (https://disc.gsfc.nasa.gov/) to download. (This data needs to be downloaded worldwide)\n
3. ERA5-Land data\n
From the European centre for medium-range weather forecasts (https://cds.climate.copernicus.eu/) to download. (This data needs to be downloaded for the range 0° E - 150° E, 81° N - 11° S)", "ds_quality": "
The matched data of AERONET sites that were not used for training were taken for verification. The mean absolute error (MAE) was 0.0515. The root mean square error (RMSE) was 0.0805. The mean deviation (MBE) was -0.0252. The correlation coefficient (R) was 0.8559. The expected error (EE) was 81.74%.", "ds_ref_way": "", "ds_abstract": "
This data set is based on the MCD19A2, MERRA-2, ERA5_land data sets. Through data pre-processing, data sampling, space-time matching, the aerosol optical thickness (AOD), wavelength index (AE), meteorological parameters and other data are extracted. The AERONET ground station observation values are taken as the target truth values, and the machine learning algorithm LightGBM training model is used to retrieve the daily 1-km resolution land dust aerosol optical thickness in the \"the Belt and Road\" desert area from 2000 to 2022. This dust aerosol optical depth dataset provides basic data for dust research in the \"the Belt and Road\" area.\n
1. Dataset Name: Dust-Aerosol-Optica-Depth\n
2. Attribute information\n
Lon: Longitude of the dataset (°)\n
Lat: Latitude of the dataset (°)\n
Dust-AOD: Optical thickness of dust aerosols
", "ds_time_res": "日", "ds_acq_place": "The Belt and Road desert region in Asia", "ds_space_res": "1km", "ds_projection": "", "ds_process_way": "1. Remote sensing data preprocessing:MERRA-2(0.5°×0.625°) and ERA5-Land(0.1°×0.1) were sampled to a resolution of 1 km by bilinear interpolation, respectively. ERA5-Land also needed to fill in the missing values by nearest neighbor interpolation before this, and then extracted the required data by spatio-temporal matching according to the nearest pixel. The aerosol optical thickness (AOD) with wavelength index (AE) <1 was selected from the AERONET ground station observation data in the study area as the target truth value of dust AOD.\n
2. Inversion method:The two models were trained respectively using the LightGBM model: (1) The complete model, that is, the model with no missing predictors, was used as the input training model (MCD19A2 data had a large number of missing data, for the non-missing part); (2) Incomplete model, that is, MCD19A2 has missing data, and only the predictors extracted from MERRA-2 and ERA5-Land data are used as input training models. The output of the two models needs to be averaged with the sand AOD data of MERRA-2 after downscaling before it can be used as the final inversion result.\n
Reference：Ke G, Meng Q, Finley T, et al. Lightgbm: A highly efficient gradient boosting decision tree[J]. Advances in neural information processing systems, 2017, 30.", "ds_ref_instruction": "" } }, "submit_center_id": "ncdc", "data_level": 0, "license_type": "CC BY 4.0", "doi_reg_from": "reg_local", "cstr_reg_from": "reg_local", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "ds_topic_tags": [ "亚洲", "陆地沙尘气溶胶" ], "ds_subject_tags": [ "自然地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "亚洲" ], "ds_time_tags": [ 2000, 2022 ], "ds_contributors": [ { "true_name": "吴小林", "email": "220220942371@lzu.edu.cn", "work_for": "兰州大学信息科学与工程学院", "country": "中国" }, { "true_name": "王兆滨", "email": "wangzhb@lzu.edu.cn", "work_for": "兰州大学", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "吴小林", "email": "220220942371@lzu.edu.cn", "work_for": "兰州大学信息科学与工程学院", "country": "中国" } ], "ds_managers": [ { "true_name": "王兆滨", "email": "wangzhb@lzu.edu.cn", "work_for": "兰州大学", "country": "中国" } ], "category": "沙漠与荒漠化" }