淮北平原Angstrom公式参数校正及太阳总辐射时空特征分析

为了解淮北平原太阳总辐射的时空特征,基于安徽省新马桥试验站内自动气象站2011-2016年的实测太阳总辐射Rs数据,分别采用最小二乘法和遗传算法对Angstrom公式中经验系数a,b值进行率定,对比分析2种率定值与FAO建议值计算ET0和Rs的差别;基于优选的率定参数,选取砀山、亳州、宿县、蚌埠及阜阳5个典型气象站点,对淮北平原太阳总辐射的时空分布特性进行分析.结果表明,FAO建议的经典参数a,b值并不适用于淮北平原,其较大估计了Rs和ET0.通过最小二乘法和遗传算法率定得出的a,b值均可有效提高Rs和ET0两者计算的准确性,且遗传算法率定的结果更优.1955-2014年的60 a内,淮北平原的日均Rs从北到南总体呈现逐渐降低的趋势,且平原北部下降幅度明显低于平原南部,1980年为主要转折年,1980年之前安徽省淮北平原年日均Rs年际下降趋势不明显,1980年之后年日均Rs年际下降趋势明显;Rs季节分布特征依次为夏季、春季、秋季、冬季,全年和各季度平均Rs在空间上呈现总体北高南低、局部有所差别的分布特征;此外,月尺度上Rs年内变化趋势呈现为单峰型分布,最大峰值出现在6月份,12月份最低.文中的研究成果可为淮北平原常规气象数据精确计算Rs和ET0提供参考.

Correction of parameters in Angstrom formula and analysis of total solar radiation characteristics in Huaibei plain

Based on the total solar radiation data observed by automatic meteorological equipment in the Xinmaqiao Experimental Station, Anhui province, during 2011-2016, the empirical coefficients a and b in the Angstrom formula were calibrated by least squares method and genetic algorithm. Additionally, ET0 and total solar radiation calculated by the calibrated values were compared with the re commended values by FAO. Based on the optimized empirical parameters and the data observed in five typical meteorological stations located in Dangshan, Bozhou, Suxian, Bengbu and Fuyang, the temporal and spatial distribution characteristics of the total solar radiation in the Huaibei plain were analyzed. The results show that the recommended values of a and b by FAO are unsuitable to the total solar radiation in the Huaibei plain because of overestimated Rs and ET0. The calibrated values by the least squares method and genetic algorithm can result in more accurate Rs and ET0. In 1955-2014, the average daily total solar radiation presents a downtrend from north to south in the plain, and the decline is larger in the north than in the south. The year of 1980 is a turning point before that adaily total solar radiation declined tendency is not obvious, but after that the radiation is declined obviously. The seasonal distribution characteristics of Rs from the largest to the smallest are: summer> spring> autumn>winter. The yearly and seasonally average Rs distributions are higher in the north and lower in the south generally, but there is a difference in Rs locally. Besides, the monthly Rs shows a unimodal distribution, its maximum is in June and the minimum is in December. These results can provide a reference for accurately calculating Rs and ET0 based on conventional meteorological data in the plain.