中国716个气象站太阳辐射日均值数据集（1961-2010）

该数据集共包含717个文件，其中station.txt文件主要描述716个站的站点信息，每列分别对应为：经度、纬度和高程；另外以站号命名的716个文件对应716个站的数据，文件中每列分别为：年、月、日和日平均太阳辐射。
该数据是基于中国气象局常规气象观测要素：温度、湿度、气压和日照时数等估算得到的。估算方法采用两个模型得到，分别为：人工神经网络模型和Yang混合模型。Yang混合模型在晴天情况下考虑了气溶胶散射和吸收、瑞利散射、水汽吸收、臭氧吸收和均一混合气体吸收五中衰减过程，云天情况下通过日照时数来参数化云对辐射的影响；而人工神经网络模型利用ANN模型在每个辐射站上建立了辐射和地面常规气象变量的关系。由于人工神经网络模型精度要比Yang混合模型估算精度高，因此通过人工神经网络模型估算值在月尺度上动态校正Yang混合模型估算值最终得到数据集合。

This dataset contains a total of 717 files. The station.txt file primarily documents the site information of 716 stations, with each column corresponding to longitude, latitude, and elevation respectively. Additionally, 716 files named after their respective station IDs hold the observation data of the 716 stations, where each column contains year, month, day, and daily average solar radiation.

This data is estimated based on conventional meteorological observation elements from the China Meteorological Administration, including temperature, humidity, atmospheric pressure, sunshine duration and other related variables. Two models are adopted for the estimation: the Artificial Neural Network (ANN) model and the Yang Hybrid Model. Under clear-sky conditions, the Yang Hybrid Model considers five attenuation processes: aerosol scattering and absorption, Rayleigh scattering, water vapor absorption, ozone absorption, and absorption by uniformly mixed gases. For cloudy-sky conditions, the impact of clouds on radiation is parameterized using sunshine duration. The ANN model establishes the correlation between solar radiation and conventional ground meteorological variables at each radiation station. Since the ANN model has higher estimation accuracy than the Yang Hybrid Model, the final dataset is obtained by dynamically correcting the estimates from the Yang Hybrid Model using the values estimated by the ANN model at the monthly scale.

搜集汇总

背景与挑战

背景概述

该数据集是中国716个气象站从1961年至2010年的太阳辐射日均值数据，基于温度、湿度等常规气象观测要素，通过人工神经网络模型和Yang混合模型混合估算生成，以提高数据精度；数据集包含站点位置信息和日平均辐射值，适用于气候研究和太阳辐射分析，并通过国家青藏高原科学数据中心平台开放共享。