本数据旨在利用公开卫星影像提取2015年与2020年中国光伏分布数据。该数据集的发布将为可再生能源、遥感、地理与环境科学等领域的研究者及用户提供重要参考。其潜在应用包括：(1)分析不同土地覆盖与土地利用类型下中国光伏设施的时空分布模式； (2) 提供可用于深度学习模型训练的中国光伏样本；(3) 估算光伏发电量及碳减排效益；(4) 评估光伏对水文与区域气候的环境影响；(5) 为未来国家及省级能源规划提供参考依据。", "ds_source": "
本研究采用Landsat卫星影像进行分析，因其拥有最长的地球表面观测记录。我们利用Google Earth Engine（GEE）平台提供的Landsat-8大气校正地表反射率产品（LANDSAT_LC08_C02_T1_L2），该产品空间分辨率为30米，并已完成大气校正与几何校正。对该产品进行了预处理操作，包括对中国全境图像的裁剪、时间滤波和除云处理。我们筛选出云量覆盖率大于10%的图像，并通过两年度各像素中位数值合成去云图像。最终分别生成2015年（1月至12月）和2020年（1月至12月）全年的中国遥感合成影像。", "ds_process_way": "
本研究采用随机森林分类器，基于谷歌地球引擎的Landsat-8影像，提取2015年与2020年中国全境光伏设施分布。通过形态学滤波、空洞填充及人工调整对结果进行可视化检验与优化。整体工作流程包括卫星图像合成、样本采集、特征提取、PV分类、后处理、精度评估、一致性评价及统计分析。", "ds_quality": "
验证分析表明，初始分类结果在2015年和2020年均实现96%以上的整体准确率。通过独立测试样本的进一步验证表明，最终数据集的精度优于现有光伏数据集。", "ds_acq_start_time": "2015-01-01 00:00:00", "ds_acq_end_time": "2020-12-31 00:00:00", "ds_acq_place": "中国", "ds_acq_lon_east": 135.01666666666668, "ds_acq_lat_south": 3.8666666666666667, "ds_acq_lon_west": 73.66666666666667, "ds_acq_lat_north": 53.5, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 8775101, "ds_files_count": 2, "ds_format": "shp", "ds_space_res": "", "ds_time_res": "年", "ds_coordinate": "WGS84", "ds_projection": "", "ds_thumbnail": "4b759a77-0cfa-4b78-9ae6-d8e2f95edd00.png", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": null, "organization_id": "0a4269e1-65f4-45f1-aeba-88ea3068eebf", "ds_serv_man": "敏玉芳", "ds_serv_phone": "09314967596", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [], "quality_level": 3, "publish_time": "2025-10-31 14:48:25", "last_updated": "2026-01-14 10:59:10", "protected": false, "protected_to": null, "lang": "zh", "cstr": "11738.11.NCDC.ZENODO.DB7003.2025", "i18n": { "en": { "title": "Chinese Vectorized Solar Photovoltaic Installed Capacity Dataset (2015-2020)", "ds_format": "shp", "ds_source": "
This study used Landsat satellite imagery for analysis, as it has the longest record of Earth surface observations. We utilize the Landsat-8 atmospheric correction surface reflectance product (LANDSAT_LC08_C02_T1_L2) provided by the Google Earth Engine (GEE) platform, which has a spatial resolution of 30 meters and has completed atmospheric and geometric correction. The product has undergone preprocessing operations, including cropping, temporal filtering, and cloud removal of images throughout China. We selected images with cloud coverage greater than 10% and synthesized cloud free images based on the median values of each pixel over two years. Finally, Chinese remote sensing composite images were generated for the entire years of 2015 (January to December) and 2020 (January to December).", "ds_quality": "
Verification analysis shows that the initial classification results achieved an overall accuracy of over 96% in both 2015 and 2020. Further verification through independent test samples shows that the accuracy of the final dataset is superior to existing photovoltaic datasets.", "ds_ref_way": "", "ds_abstract": "
This data aims to extract China's photovoltaic distribution data for 2015 and 2020 using publicly available satellite imagery. The release of this dataset will provide important references for researchers and users in fields such as renewable energy, remote sensing, geography and environmental science. Its potential applications include: (1) analyzing the spatiotemporal distribution patterns of photovoltaic facilities in China under different land cover and land use types; (2) Provide Chinese photovoltaic samples that can be used for deep learning model training; (3) Estimate the photovoltaic power generation and carbon reduction benefits; (4) Assess the environmental impact of photovoltaics on hydrology and regional climate; (5) Provide reference basis for future national and provincial energy planning.
", "ds_time_res": "年", "ds_acq_place": "China", "ds_space_res": "", "ds_projection": "", "ds_process_way": "This study used a random forest classifier based on Landsat-8 images from Google Earth Engine to extract the distribution of photovoltaic facilities throughout China in 2015 and 2020. Visualize and optimize the results through morphological filtering, hole filling, and manual adjustment. The overall workflow includes satellite image synthesis, sample collection, feature extraction, PV classification, post-processing, accuracy evaluation, consistency evaluation, and statistical analysis.", "ds_ref_instruction": "" } }, "submit_center_id": "ncdc", "data_level": 0, "license_type": "CC BY 4.0", "doi_reg_from": "reg_outside", "cstr_reg_from": "reg_outside", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "ds_topic_tags": [ "太阳能", "光伏", "可再生资源" ], "ds_subject_tags": [], "ds_class_tags": [], "ds_locus_tags": [ "中国" ], "ds_time_tags": [ 2015, 2016, 2017, 2018, 2019, 2020 ], "ds_contributors": [ { "true_name": "李龙辉", "email": "longhui.li@njnu.edu.cn", "work_for": "南京师范大学地理科学学院", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "李龙辉", "email": "longhui.li@njnu.edu.cn", "work_for": "南京师范大学地理科学学院", "country": "中国" } ], "ds_managers": [ { "true_name": "李龙辉", "email": "longhui.li@njnu.edu.cn", "work_for": "南京师范大学地理科学学院", "country": "中国" } ], "category": "其他" }