末次冰期最盛期（Last Glaciation Maximum, LGM）是距今最近的大规模冰川作用时期，这一时期以来气候演变跨越多个状态，为更好地理解多年冻土与气候系统的复杂交互提供了重要机会。然而，由于多年冻土的复杂演化过程，现有模型还缺乏可靠的模拟能力，并且分辨率较低（约100-200km），不确定性大。基于高空间分辨率的古气候数据CHELSA-TraCE21k，利用双机器学习（随机森林和XGBoost）集合模型，利用当前时期大量实测数据经过良好训练，模拟产生了22ka BP以来每100年、空间分辨率为1 km的多年冻土地温数据产品，该产品是万年尺度上目前北半球时空分辨率最高的多年冻土数据产品，对于理解末次冰期最盛期以来多年冻土动态及其与气候系统的关系等具有重要意义。\n
数据文件以GeoTiff格式提供，可通过所有常见的GeoTiff程序读取。所有数据均采用基于WGS 84水平基准的地理坐标系。每个时间段（timeID）文件夹中包含三个文件，分别表示200次模拟MAGT的集合平均值（Mean_ensemble_timeID_MAGT200）、第97.5百分位数（Quan975_ensemble_timeID_MAGT200）和第2.5百分位数（Quan025_ensemble_timeID_MAGT200），以及MAGT ≤ 0°C的次数（CountNegative_ensemble_timeID_MAGT200）。该次数除以200可用于表示多年冻土发生的概率。", "ds_source": "
", "ds_process_way": "
", "ds_quality": "
", "ds_acq_start_time": null, "ds_acq_end_time": null, "ds_acq_place": "北半球", "ds_acq_lon_east": 180.0, "ds_acq_lat_south": 25.0, "ds_acq_lon_west": -180.0, "ds_acq_lat_north": 84.0, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "open-access", "ds_total_size": 741232224963, "ds_files_count": 1335, "ds_format": "GeoTiff", "ds_space_res": "1km", "ds_time_res": "100年", "ds_coordinate": "WGS84", "ds_projection": "", "ds_thumbnail": "d23ac5fb-6c61-40c1-8552-6e932fa0ba94.jpg", "ds_thumb_from": 0, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": null, "organization_id": "52b7b79b-860c-49a5-9083-9a70cf8bed5a", "ds_serv_man": null, "ds_serv_phone": null, "ds_serv_mail": null, "doi_value": "", "subject_codes": [ "170.4510" ], "quality_level": 0, "publish_time": "2026-06-29 10:25:32", "last_updated": "2026-07-02 09:17:53", "protected": false, "protected_to": "2027-06-20 00:00:00", "lang": "zh", "cstr": "11738.11.ncdc.nieer.db7470.2026", "i18n": { "en": { "title": "Northern Hemisphere Permafrost ground temperature data since the last glacial maximum", "ds_format": "GeoTiff", "ds_source": "
", "ds_quality": "
", "ds_ref_way": "", "ds_abstract": "
The Last Glacial Maximum (LGM) is the most recent period of large-scale glaciation. Since this period, climate evolution has spanned multiple states, providing an important opportunity to better understand the complex interactions between permafrost and the climate system. However, due to the complex evolution process of permafrost, existing models still lack reliable simulation capabilities, and have low resolution (about 100-200km) and high uncertainty. Based on high spatial resolution paleoclimate data CHELSA-TraCE 21k, using dual machine learning The (Random Forest and XGBoost) ensemble model uses a large number of measured data in the current period and is well trained to simulate and produce a permafrost ground temperature data product with a spatial resolution of 1 km every 100 years since 22ka BP. This product is a permafrost data product with the highest temporal and spatial resolution in the northern hemisphere on a ten-thousand-year scale. It is of great significance for understanding the dynamics of permafrost and its relationship with the climate system since the peak of the last ice age.\r\n
Data files are provided in GeoTiff format and can be read through all common GeoTiff programs. All data use a geographical coordinate system based on the WGS84 horizontal benchmark. Each time period (timeID) folder contains three files representing the collective mean (Mean_ensemble_timeID_MAGT200), the 97.5th percentile (Quan975_ensemble_timeID_MAGT200) and the 2.5th percentile (Quan025_ensemble_timeID_MAGT200) of the 200 simulated MAGT200 times, and the number of times the MAGT200 times the MAGT200 times. This number divided by 200 can be used to express the probability of permafrost occurring.", "ds_time_res": "", "ds_acq_place": "Northern Hemisphere", "ds_space_res": "", "ds_projection": "", "ds_process_way": "
", "ds_ref_instruction": "" } }, "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": [ "冻土", "冻土分布", "古气候重建", "地温", "多年冻土", "时空演变" ], "ds_subject_tags": [ "自然地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "北半球", "北极", "北美", "欧亚大陆", "青藏高原" ], "ds_time_tags": [], "ds_contributors": [ { "true_name": "冉有华", "email": "ranyh@lzb.ac.cn", "work_for": "中国科学院西北生态环境资源研究院", "country": "中国" }, { "true_name": "李新", "email": "xinli@itpcas.ac.cn", "work_for": "中国科学院青藏高原研究所", "country": "中国" }, { "true_name": "陈发虎", "email": "fhchen@itpcas.ac.cn", "work_for": "兰州大学", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "冉有华", "email": "ranyh@lzb.ac.cn", "work_for": "中国科学院西北生态环境资源研究院", "country": "中国" } ], "ds_managers": [ { "true_name": "冉有华", "email": "ranyh@lzb.ac.cn", "work_for": "中国科学院西北生态环境资源研究院", "country": "中国" } ], "category": "冻土" }