数据集摘要：\n中国第二次冰川编目(SCGI v1)是基于 Landsat TM/ETM+、ASTER、SPOT-5 等遥感影像和 DEM 数据，采用遥感与 GIS 方法、波段比值/NDSI 阈值提取、人工修订、山脊线分割和属性重建生成的第二次中国冰川编目；SCGI v2 进一步通过 2007–2009 年 Landsat TM 少云合成影像补充原始 SCGI v1 缺失区，数据包含冰川矢量边界、GLIMS ID、面积、高程、坡度、坡向和影像时间等属性，覆盖中国西部冰川区，约 73.2°E–105.56°E、27.0°N–49.6°N，名义制图分辨率约 15–30 m，适用于现代冰川空间格局、区域统计、变化监测和编目更新。
\n数据源描述：\n中国第二次冰川编目更新版(SCGI v2)数据集的构建主要使用了多源遥感影像、已有冰川编目、DEM 数据和辅助判读资料。原始 SCGI v1 主要基于 2003—2008 年获取的 EOS/Terra ASTER 影像和 2004—2011 年获取的 Landsat TM/ETM+ 影像编制，共使用 562 景 ASTER 影像、212 景 Landsat 影像，其中包括 29 景 ETM+ 影像，以及 8 景 SPOT-5 影像。已发布的 SCGI Version 1.0 覆盖了中国西部约 85% 的冰川，但由于青藏高原东南部以及四川、云南西部地区气候湿润、云雪覆盖频繁，部分区域未能得到有效编目。为补充这些缺失区域，本研究进一步使用 2007—2009 年 Landsat TM 影像构建无云合成影像。影像筛选条件为每年第 200—330 天获取、云量小于 30%。此外，FCGI_V2 冰川边界被用作空间约束范围，ASTER GDEM v3 用于提取流域山脊线并划分冰川单元。人工修订过程中还使用了 Landsat 30 m 多光谱影像、15 m Landsat 全色波段、Landsat 热红外波段反演的地表温度、地形等高线和 Google Earth 影像作为辅助判读资料。
\n数据加工方法：\nSCGI v2 的数据加工流程包括原始编目继承、无云影像合成、自动冰川边界提取、人工修订、冰川单元划分、碎斑处理、数据合并和属性重建。在冰川边界提取阶段，基于合成影像计算 NDSI，并采用 NDSI ≥ 0.4 提取洁净冰和积雪区域。以 FCGI_V2 冰川边界外扩 1 km 缓冲区作为冰川提取的空间约束范围。初步分类结果经过 5 × 5 矩形窗口的形态学滤波，以降低椒盐噪声。随后，对初步冰川边界进行人工检查和修订。修订过程中参考 Landsat 多光谱影像、15 m 全色波段、LST、地形等高线和 Google Earth 影像，重点处理遗漏冰川、山体阴影或云影误分类，以及冰川前进或退缩导致的冰舌变化。表碛覆盖区的修订则综合考虑影像判读、冰川出水口位置和冰面湖分布等指标。冰川单元划分方面，基于 ASTER GDEM v3 提取流域山脊线，并将冰川复合体划分为单条冰川。得到独立冰川多边形后，填充小于 5 个像元，即 4500 m² 的内部空洞，并将面积小于 0.01 km² 且与邻近冰川共享边界的小碎斑合并到相邻冰川中。最后，将新提取的缺失区域冰川边界与原始 SCGI 数据集合并，形成 SCGI v2，并重新赋予标准化属性，包括 GLIMS 冰川编号、冰川面积、高程、坡度、坡向、源影像获取时间等 20 多项属性。
\n数据质量描述：\n已有研究估计 SCGI v1 总体面积不确定性为 ±3.2%，其中表碛覆盖冰川区可达 ±17.6%；本研究新增边界采用一像元缓冲区法估算，面积不确定性约为 ±12.4%。
", "ds_source": "", "ds_process_way": "", "ds_quality": "", "ds_acq_start_time": null, "ds_acq_end_time": null, "ds_acq_place": null, "ds_acq_lon_east": 105.56, "ds_acq_lat_south": 27.0, "ds_acq_lon_west": 73.2, "ds_acq_lat_north": 49.6, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 51178066, "ds_files_count": 0, "ds_format": null, "ds_space_res": "30m", "ds_time_res": "年", "ds_coordinate": "", "ds_projection": "", "ds_thumbnail": "ba86e401-9e8c-4167-b0dc-6b06146f30bb.png", "ds_thumb_from": 0, "ds_ref_way": null, "paper_ref_way": null, "ds_ref_instruction": null, "ds_from_station": null, "organization_id": null, "ds_serv_man": null, "ds_serv_phone": null, "ds_serv_mail": null, "doi_value": "", "subject_codes": [ "170", "170.45", "170.4510" ], "quality_level": 0, "publish_time": "2026-08-05 12:06:58", "last_updated": "2026-08-05 12:18:52", "protected": false, "protected_to": "2027-01-31 00:00:00", "lang": "zh", "cstr": null, "i18n": { "en": { "title": "The Second Chinese Glacier Inventory (Version 2)", "ds_abstract": "Dataset Abstract\nThe Second Chinese Glacier Inventory (SCGI v1) was established using multi-source remote sensing imagery (Landsat TM/ETM+, ASTER, SPOT-5) and digital elevation models (DEMs) through band-ratioing, Normalized Difference Snow Index (NDSI) thresholding, manual editing, ridge-based drainage basin segmentation, and attribute assignment. SCGI v2 builds upon this foundation by incorporating cloud-reduced Landsat TM composite imagery (2007–2009) to map previously uninventoried regions. The dataset delivers vector glacier boundaries populated with standardized attributes, including GLIMS IDs, area, elevation metrics, slope, aspect, and acquisition dates. Spanning glacierized regions across western China (73.2°E–105.56°E, 27.0°N–49.6°N) at a nominal spatial resolution of 15–30 m, SCGI v2 provides a robust baseline for evaluating modern spatial glacier distribution, conducting regional inventory statistics, monitoring cryospheric changes, and updating glacier inventories.\n\nData Sources\nThe construction of SCGI v2 integrated multi-source satellite observations, legacy glacier inventories, DEMs, and auxiliary interpretation tools. The baseline SCGI v1 relied primarily on 562 EOS/Terra ASTER scenes (2003–2008), 212 Landsat TM/ETM+ scenes (2004–2011, including 29 ETM+ scenes), and 8 SPOT-5 scenes. Although SCGI v1 mapped ~85% of glaciers across western China, persistent cloud cover and seasonal snowpack in humid zones—specifically the southeastern Tibetan Plateau and western Sichuan and Yunnan provinces—prevented complete inventorying.\n\nTo reconcile these data gaps, we generated cloud-reduced composite imagery from Landsat TM scenes acquired between 2007 and 2009, filtered for Day-of-Year (DOY) 200–330 with cloud cover <30%. Additionally, glacier outlines from the First Chinese Glacier Inventory (FCGI_V2) defined the spatial constraints for extraction, while ASTER GDEM v3 was used to extract watershed ridgelines and delineate individual glacier units. Auxiliary data for visual interpretation and manual correction included 30-m Landsat multispectral imagery, 15-m panchromatic bands, Landsat-derived land surface temperature (LST) fields, topographic contours, and high-resolution Google Earth imagery.\n\nData Processing Methodology\nThe SCGI v2 processing pipeline encompasses legacy data integration, cloud-free image compositing, automated glacier extraction, manual refinement, ice-divide segmentation, topological cleaning, data merging, and attribute recalculation:\n\nAutomated Glacier Extraction: Clean ice and snow cover were classified using an NDSI threshold (NDSI ≥ 0.4) applied to the composite imagery, constrained within a 1-km outer buffer around the FCGI_V2 outlines. Morphological filtering with a 5 × 5 pixel rectangular window was applied to suppress salt-and-pepper noise.\n\nManual Refinement: Initial boundaries were manually inspected and corrected against Landsat multispectral/panchromatic images, LST fields, topographic contours, and Google Earth imagery. This phase prioritized resolving unmapped glaciers, terrain/cloud shadow misclassifications, and terminus position changes resulting from retreat or advance. Debris-covered ice was delineated by jointly assessing spectral signatures, proglacial meltwater stream outlets, and supraglacial lake distributions.\n\nGlacier Unit Segmentation & Cleaning: Glacier complexes were partitioned into discrete glacier entities along drainage divides derived from ASTER GDEM v3. Internal polygon voids smaller than 5 pixels (4,500 m²) were filled. Isolated slivers (<0.01 km²) sharing boundaries with adjacent glaciers were merged into the larger ice body.\n\nData Merging & Attribute Recalculation: The newly extracted gap-fill outlines were merged with the baseline SCGI v1 dataset to form SCGI v2. Each glacier entity was assigned standardized GLIMS attributes, encompassing over 20 topographic, spatial, and temporal parameters (e.g., GLIMS ID, glacier area, mean/min/max elevation, slope, aspect, and source image acquisition date).\n\nData Quality & Uncertainty\nPrevious evaluations estimated the overall area uncertainty of SCGI v1 at ±3.2%, rising to ±17.6% in debris-covered sectors. For the newly mapped glacier outlines introduced in SCGI v2, spatial error estimation via the standard one-pixel buffer method yields an area uncertainty of approximately ±12.4%.", "ds_time_res": "years", "ds_space_res": "30", "ds_projection": "" } }, "submit_center_id": "ncdc", "data_level": 0, "recommendation_value": 0, "license_type": "https://creativecommons.org/licenses/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, "belong_to_nieer": false, "allow_update_data": false, "ds_topic_tags": [ "第二次中国冰川编目", "冰川", "中国西部" ], "ds_subject_tags": [ "地球科学", "地理学", "自然地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "中国西部", "青藏高原" ], "ds_time_tags": [ 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013 ], "ds_contributors": [ { "true_name": "谢福明", "email": "fuming.xie@ynu.edu.cn", "work_for": "云南大学 地球科学学院", "country": "中国" }, { "true_name": "刘时银", "email": "shiyin.liu@ynu.edu,cn", "work_for": "云南大学", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "谢福明", "email": "fuming.xie@ynu.edu.cn", "work_for": "云南大学 地球科学学院", "country": "中国" }, { "true_name": "刘时银", "email": "shiyin.liu@ynu.edu,cn", "work_for": "云南大学", "country": "中国" } ], "ds_managers": [ { "true_name": "谢福明", "email": "fuming.xie@ynu.edu.cn", "work_for": "云南大学 地球科学学院", "country": "中国" } ], "category": "冰川" }