基于云模型的西北地区辐射时空分布研究

Characteristics of Temporal-spatial Solar Radiation Distribution in Northwest China Based on Cloud Model

为研究西北地区辐射(Rs)时空分布特性,基于西北地区16个辐射站点4个分区(Ⅰ区,新疆北部、甘肃河西走廊中西段、宁夏中北部、内蒙古西部;Ⅱ区,新疆南部;Ⅲ区,青海省;Ⅳ区,甘肃东南部、宁夏南部、陕西关中地区与陕北地区) 1995—2015年实测日Rs与日照时数(n)率定ngstr9m-Prescott(A-P)模型参数a、b,使用A-P模型4种不同参数率定方法(M1:分月率定,M2:分季率定,M3:分半年率定,M4:多年率定),选取其中精度高且简便的方法计算4个分区1961—2015年共55 a的Rs,并用云模型描述西北地区Rs时空分布特性,结果表明:各分区4种方法计算的Rs值与实测值在日尺度和月尺度上拟合结果均较好。各分区4种方法计算Rs的RMSE、nRMSE分布相似,MBE略有差异,t检验结果表明Ⅰ区和Ⅲ区使用M1、M2、M3计算的Rs值与实测值无明显差异,Ⅱ区使用M1、M3计算的Rs值与实测值无明显差异,Ⅳ区使用M1、M2计算的Rs值与实测值无明显差异。Ⅰ区Rs时间分布的不均匀性较小但不稳定,Ⅱ区、Ⅲ区Rs时间分布的不均匀性较小且较稳定,Ⅳ区Rs时间分布的不均匀性较大且不稳定。西北地区Rs空间分布不均匀,4个季节均表现为Ⅲ区(青海省) Rs较大;与Rs在时间上的分布特性相比,其在空间上的分布特性更不均匀、不稳定。该研究结果可用于构建完整的西北地区Rs时间序列,为西北地区Rs时空变化研究提供科学参考。

In order to study the temporal-spatial distribution of solar radiation (Rs) in Northwest China, daily Rs of 16 stations in Northwest China from 1961 to 2015 were calculated by Angstrom-Prescott (A-P) model, and cloud model was used to describe the temporal-spatial distribution of Rs in Northwest China. To obtain the best parameters calibrated method and assess its adaptability, Northwest China was divided into four areas.Ⅰ area included northern Xinjiang, middle and west section of Hexi Corridor of Gansu, central and northern Ningxia, western Inner Mongolia.Ⅱ area included southern Xinjiang.Ⅲ area included Qinghai.Ⅳ area included southeast Gansu, southern Ningxia, Guanzhong and northern Shaanxi. Daily climatic data radiation (Rs) and sunshine hour (n) of 16 stations from 1995 to 2015 were used to calibrate A-P model parameters, and four different parameters calibrated methods for A P model were used, including monthly (M1), seasonally (M2), bi-annualy (M3) and annually (M4), in which high precision and simple calculation method for each area was finally selected to calculate Rs . The results showed that the fitting results of four parameters calibrated methods conformed the measured values well both in daily and monthly scales. RMSE and nRMSE of four parameters calibrated methods changed little. The values of t -statistic showed that there was no difference among the three methods (M1, M2 and M3) in use of data calculation in Ⅰ area and Ⅲ area, no difference between the two methods (M1 and M3) in use of data calculation in Ⅱ area, no difference between the two methods (M1 and M2) in use of data calculation in Ⅳ area. When cloud model was used to analyze the temporal-spatial distribution characteristics of Rs in Northwest China, as a consequence, there was good homogeneity but poor stability of Rs in Ⅰ area in the time. There was good homogeneity and stability of Rs in Ⅱ area and Ⅲ area in the time. There was poor homogeneity and stability of Rs in Ⅳ area in the time. Besides, spatial distribution of Rs in Northwest China was uneven. Rs of the four seasons in Ⅲ area (Qinghai Province) was larger than those of other areas. In general, the homogeneity and stability of R s in the space was worse than that in the time. The results can be used to construct a complete Rs time series in Northwest China and provide a scientific reference for the study of temporal-spatial distribution characteristics of Rs in Northwest China.