Global climate change intensifies the hydrological cycle, leading to frequent extreme precipitation events. The precipitation concentration index is a key tool for diagnosing the spatiotemporal distribution characteristics of precipitation. The existing research has problems such as fragmented site observations and a disconnect between historical benchmarks and future estimates. This dataset integrates ground observation and grid observation data from China from 1961 to 2022, as well as statistically downscaled CMIP6 estimation data from four SSP scenarios from 2015 to 2100, with a spatial resolution of 0.25 °. It constructs a spatiotemporal continuous dataset (MPCID) that includes four core indicators: precipitation concentration (PCD), precipitation concentration period (PCP), daily precipitation concentration index (DPCI), and monthly precipitation concentration index (MPCI).
Through site data verification, PCD has the smallest error and the best correlation. The dataset can support research on the spatiotemporal distribution of precipitation in China, hydrological and agricultural climate impact assessment, and adaptive management strategy formulation.
1. Daily surface precipitation data (1961-2020): China Meteorological Administration's daily surface climate data set (V3.0), after quality control and uniformity testing, 651 stations were selected;
2. Grid point observation data (1961-2022): CN05.1 grid point precipitation data set, 0.25 ° × 0.25 °, issued by the Climate Change Research Center of the Chinese Academy of Sciences;
3. Climate model data: 24 CMIP6 global climate models output precipitation data, covering four scenarios: SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5.
1. Data preprocessing: Quality control and missing value imputation of site data; Perform unified resolution resampling and calendar format conversion on CMIP6 mode data;
2. Spatial interpolation: KD Tree+inverse distance weight (IDW) is used to interpolate station data into grid data;
3. Statistical downscaling: Quantile mapping (QM) and quantile incremental mapping (QDM) statistical downscaling methods are used to correct bias in CMIP6 data. The optimal framework is to use quantile incremental mapping after resampling;
4. Indicator calculation: Calculate PCD and PCP based on vector analysis, calculate DPCI based on Lorenz curve, and calculate MPCI based on variance method;
5. Accuracy verification: Multi dimensional verification is conducted using indicators such as MAE, RMSE, BIAS, CORR, IVS, TS, etc.
1. Historical verification: PCD has the best accuracy with MAE=0.034, RMSE=0.042, CORR=0.902; PCP has high correlation but significant error; DPCI error is controllable, but the correlation on a daily scale is limited; MPCI has low sensitivity to extreme precipitation;
2. Future scenario verification: Under four SSP scenarios, the MAE of PCD is 0.094-0.096, and the RMSE is 0.122-0.125, indicating stable performance; The resample QDM downscaling method has the smallest error;
3. Spatial validation: The dataset can accurately depict the spatial differentiation characteristics of precipitation concentration in China, such as "high in the north and low in the west". The results of complex terrain areas need to be interpreted carefully in conjunction with station density.
| # | number | name | type |
| 1 | 42261026 | National Natural Science Foundation of China | |
| 2 | 41761014 | National Natural Science Foundation of China | |
| 3 | XJYS0907-2023-01 | other | |
| 4 | 25JR6KA005 | other |
This work is licensed under
CC BY 4.0 (Creative Commons Attribution 4.0 International License).
CSTR:11738.11.NCDC.PRECIPITATION.DB7311.2026
10.12072/ncdc.precipitation.db7311.2026
