HRLT：中国高分辨率（逐日，1km）长时序气温和降水网格数据集（1961-2019年）

高时空分辨率的精确长期气温和降水估算对于各种气候学研究至关重要。研究团队研制了一个新的、可公开获取的中国日网格最高气温、最低气温和降水量数据集，其空间分辨率高达 1 千米，时间跨度为1961-2019 年。该数据集是基于中国气象局的 0.5°× 0.5°网格数据集，以及海拔、高差、坡度、地形湿润指数、纬度和经度等协变量，利用综合统计分析对日网格数据进行了插值，包括机器学习方法、广义加法模型和薄板样条法；通过交叉验证选择最佳的集成模型，并使用薄板样条法校正残差误差，得到最终模型的表面结果；使用中国气象站点观测数据评估栅格化数据的准确性，包括中国气象局陆地数据同化系统 (CLDAS)、中国气象强迫数据集 (CMFD) 和 ISIMIP3a 历史数据集等。数据格式为NetCDF。

Accurate long-term air temperature and precipitation estimates with high spatiotemporal resolution are critical for a wide range of climatological research. A new, publicly accessible daily gridded dataset of maximum air temperature, minimum air temperature, and precipitation over China was developed by the research team, with a spatial resolution of up to 1 km and a temporal span from 1961 to 2019. This dataset is constructed based on the 0.5°×0.5° gridded dataset from the China Meteorological Administration (CMA), alongside covariates including elevation, elevation difference, slope, topographic wetness index, latitude, and longitude. Comprehensive statistical analyses, including machine learning methods, generalized additive models (GAMs), and thin plate splines, were utilized to interpolate the daily gridded data. The optimal ensemble model was selected through cross-validation, and residual errors were corrected using thin plate splines to produce the surface outputs of the final model. The accuracy of the gridded data was evaluated using in-situ meteorological station observations from China, along with reference datasets such as the China Meteorological Administration Land Data Assimilation System (CLDAS), China Meteorological Forcing Dataset (CMFD), and ISIMIP3a historical dataset. The data are stored in NetCDF format.

搜集汇总

背景与挑战

背景概述

该数据集是中国全境1961-2019年逐日最高气温、最低气温和降水量的高分辨率（1公里）网格数据，基于中国气象局0.5°网格数据，通过机器学习、广义加法模型和薄板样条法等综合统计方法插值生成，并经过交叉验证和残差校正确保精度。数据质量评估显示气温数据精度高（如最高气温平均绝对误差1.07°C，相关系数0.99），降水数据可靠性好（平均绝对误差1.30 mm，相关系数0.84），相比其他数据集具有更高空间分辨率或相当精度，适用于长期气候分析和研究。