降水时序数据是全球和区域气候研究、水文模拟及生态系统评估中的关键基础数据。然而，现有市县未来降水数据序列普遍较短。这导致当前对区域尺度，尤其是市县等局部单元上的长期降水动态与趋势变化的评估仍存在不确定性。本研究基于ACCESS-ESM1-5、IPSL-CM6A-LR、MIROC-ES2L等全球气候模式的未来情景数据，生成黄土高原不同市县（如延安市、兰州市、天水市等）2015-2100年未来降水数据产品。本数据集的核心优势在于本数据集的核心优势在于提供了2015-2100年共86年的长序列降水数据。这一完整的长时间序列有助于揭示局地气候对全球变化的响应细节，从而为市县尺度的长期气候风险评估、适应性规划提供坚实的数据基础。
\n本数据以Excel格式（*.xlsx）存储。数据文件中不同工作表为不同市县未来降水数据，ACCESS-ESM1-5、IPSL-CM6A-LR、MIROC-ES2L为不同模型，ssp126、ssp245、ssp585为排放情景，时间分辨率为年，时间范围为2015-2100年。", "ds_source": "
高分辨率观测数据采用基于欧洲中期天气预报中心第五代再分析数据（ERA5）开发的MSWX格点化偏差校正气象数据集，该数据集提供了10个近地表气象变量，空间分辨率为0.1°，时间分辨率为3小时，作为模型训练和验证的对照组（https://www.gloh2o.org/mswx/）。基础气候数据来自耦合模式比较计划第六阶段（CMIP6）中3个全球气候模式的输出，包括历史时期（1950-2014年）和三种未来情景（SSP1-2.6, SSP2-4.5, SSP5-8.5，2015-2100年）的逐日模拟，其空间分辨率介于0.7°至2.8°之间（https://esgf-node.llnl.gov/search/cmip6/）
", "ds_process_way": "（1）利用MSWX数据集作为高分辨率观测数据，用于后续偏差校正与模型训练；（2）采用分位数Delta映射方法对CMIP6的3个全球气候模式输出进行偏差校正，校正后的数据用于降尺度输入；（3）基于U-Net深度学习架构构建降尺度模型，针对不同季节分别训练四个独立的U-Net模型，输入包括地形数据与偏差校正后的气象变量，并采用改进的损失函数以提升极端事件模拟能力；（4）利用1982–2014年期间的数据进行模型训练与验证，通过对比MSWX观测数据，从气候均值、极端事件频率及未来变化等多个方面系统评估降尺度数据集的性能。（5）基于经验证可靠的降尺度格点数据集，利用GIS空间分析技术依据标准的市县级行政区矢量边界，计算每个行政单元内所有格点的年度降水量平均值，生成黄土高原不同市县降水的逐年序列。
", "ds_quality": "本研究采用均方根误差（RMSE）、平均偏差（MB）以及对不同等级降水事件（小雨、中雨、大雨、暴雨）频率的刻画能力作为核心验证指标，系统评估了降尺度生成的降水数据与MSWX高分辨率观测数据在1979年至2014年整个历史时期的一致性。
\n验证结果表明，降尺度后数据的精度显著提升：大部分CMIP6模式的降水RMSE从原始输出的约1.00毫米/天降至0.37毫米/天以下；空间上，多模式集合平均的偏差在黄土高原地区被控制在-0.6至0.6毫米/天之间，有效纠正了原始数据普遍存在的系统性高估。因此，该降尺度降水数据集能够更可靠地反映黄土高原区域不同市县降水的实际情况，可作为黄土高原不同市县水文气候研究的可靠数据基础。
", "ds_acq_start_time": "2015-01-01 00:00:00", "ds_acq_end_time": "2100-12-31 00:00:00", "ds_acq_place": "黄土高原", "ds_acq_lon_east": 114.55, "ds_acq_lat_south": 33.75, "ds_acq_lon_west": 100.95, "ds_acq_lat_north": 41.25, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 200468, "ds_files_count": 0, "ds_format": "*.xlsx", "ds_space_res": "", "ds_time_res": "", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "407a9e6d-e0cb-4580-8760-9ec95d5378f7.png", "ds_thumb_from": 0, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": "", "organization_id": "bf138922-7121-438c-8d1b-19d5f751c907", "ds_serv_man": "李红星", "ds_serv_phone": "0931-4967592", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [ "170.15", "170.4510" ], "quality_level": 0, "publish_time": "2026-05-21 10:09:45", "last_updated": "2026-05-21 10:51:05", "protected": false, "protected_to": null, "lang": "zh", "cstr": "11738.11.ncdc.loess.db7334.2026", "i18n": { "en": { "title": "Annual Precipitation Dataset for Different Cities and Counties on the Loess Plateau (2015–2100)", "ds_format": "*.xlsx", "ds_source": "The high-resolution observational data used in this study are based on the MSWX gridded bias-corrected meteorological dataset developed from the European Centre for Medium-Range Weather Forecasts (ECMWF) fifth-generation reanalysis (ERA5). This dataset provides 10 near-surface meteorological variables at a spatial resolution of 0.1° and a temporal resolution of 3 hours, serving as the \"ground truth\" for model training and validation (https://www.gloh2o.org/mswx/). The baseline climate data were derived from outputs of three global climate models participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6), including historical simulations (1950–2014) and daily projections under three future scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5, 2015–2100). The spatial resolution of these climate models ranges from 0.7° to 2.8° (https://esgf-node.llnl.gov/search/cmip6/).", "ds_quality": "
This study employed Root Mean Square Error (RMSE), Mean Bias (MB), and the capability to characterize the frequency of different categories of precipitation events (light rain, moderate rain, heavy rain, and rainstorm) as the core validation metrics. These were used to systematically evaluate the consistency between the downscaled precipitation data and the high-resolution MSWX observational dataset over the entire historical period from 1979 to 2014.The validation results demonstrate a significant improvement in the accuracy of the downscaled data. For most CMIP6 models, the precipitation RMSE was reduced from approximately 1.00 mm/day in the raw model outputs to below 0.37 mm/day. Spatially, the multi-model ensemble mean bias was controlled within the range of -0.6 to 0.6 mm/day over the Loess Plateau region, effectively correcting the systematic overestimation commonly present in the original data. Therefore, this downscaled precipitation dataset can more reliably reflect the actual precipitation conditions across different cities and counties within the Loess Plateau region. It serves as a reliable data foundation for hydroclimatic research concerning various cities and counties on the Loess Plateau.", "ds_ref_way": "", "ds_abstract": "
Precipitation time series data are critical foundational data for global and regional climate research, hydrological modeling, and ecosystem assessments. However, existing future precipitation data series for cities and counties are generally short in duration. This introduces uncertainty in assessing long-term precipitation dynamics and trends at regional scales, particularly for local administrative units such as cities and counties.This study utilizes future precipitation datasets from ACCESS-ESM1-5, IPSL-CM6A-LR, and MIROC-ES2L. A downscaling model based on the U-Net deep learning architecture was constructed to generate a future precipitation data product for the Loess Plateau covering the period from 2015 to 2100. The spatial coverage of the dataset spans from 33.75°N to 41.25°N in latitude and 100.95°E to 114.55°E in longitude.The core advantage of this dataset lies in its provision of a long-term, 86-year precipitation data series from 2015 to 2100. This complete, extended temporal record facilitates the detailed analysis of local climate responses to global change, thereby providing a solid data foundation for long-term climate risk assessment and adaptation planning at the city and county scale.\r\n
The dataset is stored in Excel format (*.xlsx). Different worksheets correspond to future precipitation data of different cities and counties. ACCESS-ESM1-5, IPSL-CM6A-LR and MIROC-ES2L represent different climate models, while ssp126, ssp245 and ssp585 denote different emission scenarios. The dataset has an annual temporal resolution and covers the period from 2015 to 2100.", "ds_time_res": "", "ds_acq_place": "Loess Plateau", "ds_space_res": "", "ds_projection": "", "ds_process_way": "
(1) The MSWX dataset was utilized as high-resolution observational data for subsequent bias correction and model training. (2) The Quantile Delta Mapping method was applied to perform bias correction on the outputs from three CMIP6 global climate models, with the corrected data then serving as input for downscaling. (3) A downscaling model was constructed based on the U-Net deep learning architecture. Four independent U-Net models were trained for different seasons, with inputs including terrain data and bias-corrected meteorological variables. A modified loss function was employed to enhance the simulation capability for extreme events. (4) Data from the period 1982–2014 were used for model training and validation. The performance of the downscaled dataset was systematically evaluated by comparing it with MSWX observational data across multiple aspects, including climatic means, extreme event frequency, and future changes. (5) Based on the validated and reliable downscaled gridded dataset, GIS spatial analysis techniques were applied in accordance with standard city/county-level administrative vector boundaries. The annual average precipitation value for all grid points within each administrative unit was calculated, generating annual time series of precipitation for different cities and counties on the Loess Plateau.", "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, "ds_topic_tags": [ "降水", "长时序", "深度学习" ], "ds_subject_tags": [ "大气科学", "自然地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "延安市", "兰州市", "天水市", "庆阳市", "西吉县", "泾阳县", "北吴起县", "绥德县", "衡山县" ], "ds_time_tags": [ 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026, 2027, 2028, 2029, 2030, 2031, 2032, 2033, 2034, 2035, 2036, 2037, 2038, 2039, 2040, 2041, 2042, 2043, 2044, 2045, 2046, 2047, 2048, 2049, 2050, 2051, 2052, 2053, 2054, 2055, 2056, 2057, 2058, 2059, 2060, 2061, 2062, 2063, 2064, 2065, 2066, 2067, 2068, 2069, 2070, 2071, 2072, 2073, 2074, 2075, 2076, 2077, 2078, 2079, 2080, 2081, 2082, 2083, 2084, 2085, 2086, 2087, 2088, 2089, 2090, 2091, 2092, 2093, 2094, 2095, 2096, 2097, 2098, 2099, 2100 ], "ds_contributors": [ { "true_name": "张宝庆", "email": "baoqzhang@lzu.edu.cn", "work_for": "兰州大学", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "张宝庆", "email": "baoqzhang@lzu.edu.cn", "work_for": "兰州大学", "country": "中国" } ], "ds_managers": [ { "true_name": "张宝庆", "email": "baoqzhang@lzu.edu.cn", "work_for": "兰州大学", "country": "中国" } ], "category": "气象" }