本数据集为全球月度0.5°×0.5°过火面积比例数据集，覆盖时段为1901年至2020年。数据基于气候条件、植被状态、人口密度和土地利用数据，通过机器学习模型重建生成，旨在提供20世纪时空一致的过火面积数据，以弥补卫星观测时间覆盖短和模型模拟不确定性大的不足。数据集包含三个版本：基于FireCCI51卫星产品训练的版本、基于GFED5卫星产品训练的版本，以及用历史统计过火面积与GDP关系校准的FireCCI51-GDP版本。数据变量包括网格中心纬度、经度、过火面积比例以及区分常规与极端火灾的类型标识。", "ds_source": "
数据来源于https://zenodo.org/records/14191467 。", "ds_process_way": "
数据生产首先将全球划分为14个GFED区域并分别处理。采用分类模型以区域内过火面积比例的90%分位数为阈值，将网格单元识别为极端或常规火灾。随后，针对两类火灾分别训练回归模型，模型以2003-2020年期间排除农田火灾的卫星过火面积产品（FireCCI51）为训练目标，利用气候、植被、人口与土地利用作为预测变量。训练后的模型应用于1901-2020年的历史数据以重建过火面积。此外，还生产了以GFED5为训练目标的版本，以及对FireCCI51版本利用历史统计过火面积与GDP关系进行区域尺度校准的版本。", "ds_quality": "
数据集通过未参与训练的卫星观测数据、木炭记录以及其他独立全球和区域过火面积数据集进行了验证。其优势在于提供了20世纪长期、空间一致的过火面积重建数据，且通过分区建模和区分极端与常规火灾的方法提升了表征能力。局限性包括：百年尺度重建依赖于模型外推，其不确定性受历史输入数据精度影响；训练所用的卫星产品本身存在不确定性且排除了农田火灾；GDP校准版本建立在历史关系假设之上；早期历史时期缺乏直接观测验证；空间分辨率较粗。该数据集主要用于评估动态全球植被模型中火灾模块的模拟结果。", "ds_acq_start_time": "1901-01-01 00:00:00", "ds_acq_end_time": "2020-12-31 00:00:00", "ds_acq_place": "全球", "ds_acq_lon_east": 180.0, "ds_acq_lat_south": -90.0, "ds_acq_lon_west": -180.0, "ds_acq_lat_north": 90.0, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 5296954588, "ds_files_count": 342, "ds_format": "*.nc", "ds_space_res": "", "ds_time_res": "月", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "7ff2361c-fdc8-4a05-9f53-8906d07aeb8c.png", "ds_thumb_from": 0, "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": "0931-4967596", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [ "170.45" ], "quality_level": 3, "publish_time": "2025-12-29 10:19:43", "last_updated": "2025-12-29 10:19:43", "protected": false, "protected_to": null, "lang": "zh", "cstr": "11738.11.NCDC.ZENODO.DB7048.2025", "i18n": { "en": { "title": "Reconstructed Global Monthly Burned Area Maps from 1901 to 2020", "ds_format": "*.nc", "ds_source": "
The data is sourced from https://zenodo.org/records/14191467 .", "ds_quality": "
The dataset was validated using satellite observation data, charcoal records, and other independent global and regional datasets of burned areas that did not participate in training. Its advantage lies in providing long-term, spatially consistent data for the reconstruction of burned areas in the 20th century, and improving its representation ability through zoning modeling and distinguishing between extreme and conventional fires. Limitations include: century scale reconstruction relies on model extrapolation, and its uncertainty is affected by the accuracy of historical input data; The satellite products used for training have inherent uncertainty and exclude the possibility of agricultural fires; The GDP calibration version is based on the assumption of historical relationships; Lack of direct observation and verification in the early historical period; The spatial resolution is relatively coarse. This dataset is mainly used to evaluate the simulation results of the fire module in the dynamic global vegetation model.", "ds_ref_way": "", "ds_abstract": "
This dataset is a global monthly 0.5 °× 0.5 ° fire area ratio dataset, covering the period from 1901 to 2020. The data is based on climate conditions, vegetation status, population density, and land use data, reconstructed and generated through machine learning models, aiming to provide spatially and temporally consistent burned area data in the 20th century to compensate for the shortcomings of short satellite observation time coverage and high model simulation uncertainty. The dataset consists of three versions: a version trained on FireCCI51 satellite products, a version trained on GFED5 satellite products, and a FireCCI51-GDP version calibrated using historical statistics on the relationship between burned area and GDP. The data variables include the latitude and longitude of the grid center, the proportion of burned areas, and the type identification that distinguishes between conventional and extreme fires.
", "ds_time_res": "月", "ds_acq_place": "global", "ds_space_res": "", "ds_projection": "", "ds_process_way": "Data production first divides the world into 14 GFED regions and processes them separately. Using a classification model with a threshold of the 90th percentile of the burned area ratio within the region, identify grid cells as extreme or conventional fires. Subsequently, regression models were trained separately for two types of fires, with the satellite burnt area product (FireCCI51) excluding farmland fires from the period of 2003-2020 as the training objective, and climate, vegetation, population, and land use as predictive variables. The trained model is applied to historical data from 1901 to 2020 to reconstruct the burned area. In addition, a version with GFED5 as the training target and a version with regional scale calibration using historical statistics on the relationship between burned area and GDP were also produced for the FireCCI51 version.", "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_outside", "cstr_reg_from": "reg_outside", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "ds_topic_tags": [ "过火面积", "机器学习重建", "火灾" ], "ds_subject_tags": [ "地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "全球" ], "ds_time_tags": [ 1901, 1902, 1903, 1904, 1905, 1906, 1907, 1908, 1909, 1910, 1911, 1912, 1913, 1914, 1915, 1916, 1917, 1918, 1919, 1920, 1921, 1922, 1923, 1924, 1925, 1926, 1927, 1928, 1929, 1930, 1931, 1932, 1933, 1934, 1935, 1936, 1937, 1938, 1939, 1940, 1941, 1942, 1943, 1944, 1945, 1946, 1947, 1948, 1949, 1950, 1951, 1952, 1953, 1954, 1955, 1956, 1957, 1958, 1959, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979, 1980, 1981, 1982, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020 ], "ds_contributors": [ { "true_name": "李伟", "email": "wli2019@tsinghua.edu.cn", "work_for": "清华大学", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "李伟", "email": "wli2019@tsinghua.edu.cn", "work_for": "清华大学", "country": "中国" } ], "ds_managers": [ { "true_name": "李伟", "email": "wli2019@tsinghua.edu.cn", "work_for": "清华大学", "country": "中国" } ], "category": "生态" }