Precipitation time series data are fundamental and crucial for global and regional climate research, hydrological modeling, and ecosystem assessments. However, existing 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. To address this, the present study utilizes version 1.7 of the China Meteorological Forcing Dataset (CMFD) to generate a historical precipitation data product for the Loess Plateau covering the period from 1982 to 2020. Historical precipitation data for various cities and counties were then extracted from this product. 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. Compared to other datasets, the core advantage of this product lies in its provision of a long-term precipitation data series spanning 39 years from 1982 to 2020. 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.
The data are stored in Excel format (*.xlsx). Different worksheets in the file contain precipitation data of different cities and counties, with annual temporal resolution covering the period from 1982 to 2020.
The China Meteorological Forcing Dataset (CMFD) version 1.7 is a high spatiotemporal resolution gridded meteorological dataset specifically developed for land surface process studies over China. It was created by the Institute of Tibetan Plateau Research, Chinese Academy of Sciences. By integrating ground-based observational data from approximately 700 stations of the China Meteorological Administration, multiple satellite remote sensing products, and international reanalysis datasets, it generates a long-term, spatiotemporally continuous near-surface meteorological forcing field covering China since 1979. CMFD version 1.7 provides seven key variables, including 2-meter air temperature, surface pressure, specific humidity, 10-meter wind speed, downward shortwave and longwave radiation, and precipitation rate. It features a high spatial resolution of 0.1° (approximately 10 kilometers) and a temporal resolution of 3 hours. Due to its high quality and excellent performance over China, it has become a crucial benchmark dataset for driving various land surface, hydrological, and ecological models, as well as for conducting regional climate analysis.
(1) The 3-hourly precipitation rate gridded data from the nationwide China Meteorological Forcing Dataset (CMFD) version 1.7 was adopted as the foundational input data source. A spatiotemporal data subset for various cities and counties within the Loess Plateau region (33.75°N–41.25°N, 100.95°E–114.55°E) was extracted and cropped from this source. (2) The high-temporal-resolution gridded data was aggregated to a daily scale. The specific method involved converting each of the eight 3-hourly precipitation rate (mm/hr) values within a day into period-specific precipitation amounts by multiplying by the time step length (3 hours), followed by summation to ultimately generate the daily total precipitation (mm/day) for each grid point. (3) Based on the obtained daily accumulated precipitation gridded data, annual totals were computed for each grid point by summing across the time dimension year by year. Subsequently, the average value of all grid points within each city/county-level administrative unit was calculated, ultimately generating an annual average precipitation time series for each city and county. (4) The accuracy and reliability of the generated annual precipitation dataset for different cities and counties on the Loess Plateau were systematically validated from multiple aspects, including Mean Bias Error (MBE), Root Mean Square Error (RMSE), and the correlation coefficient (R²), by comparing it with observed annual precipitation data from national meteorological stations within the region.
This study employed Mean Bias Error (MBE), Root Mean Square Error (RMSE), and the correlation coefficient (R²) as the core validation metrics to systematically evaluate the consistency between this precipitation dataset and independent station observations.
The validation results indicate a high degree of consistency between this precipitation data and station observations in the Loess Plateau region. The dataset exhibits a Mean Bias Error (MBE) ranging from 16 to 22 mm/year, a Root Mean Square Error (RMSE) between 11 and 17 mm/year, and a correlation coefficient (R²) from 0.92 to 0.99 when compared against the station data. With its reliable performance in terms of systematic bias, error magnitude, and correlation, this precipitation data can more accurately characterize the temporal distribution of precipitation in different cities and counties of the Loess Plateau. It is, therefore, well-suited for long-term climate risk assessment at the city and county scale.
This work is licensed under
CC BY 4.0 (Creative Commons Attribution 4.0 International License).
CSTR:11738.11.ncdc.loess.db7332.2026
10.12072/ncdc.loess.db7332.2026
