摘要
文章通过对安徽省内20个代表台站1961-2005年月平均气温序列均一性检验的研究,提出了一种新的基于Γ分布气温序列非均一性检验方法——Γ检验法;以合肥为例给出了数据处理和结果分析的全过程,并将该方法检验结果与常见的基于正态分布的SNHT法的检验结果进行了对比分析,结果表明该方法具有更广泛的使用范围和较高的准确性,可以检验大部分气温序列的非均一性;同时说明安徽省内大多数台站的温度资料质量基本可靠,而台站迁移则是造成气温序列非均一性的主要原因。
A new Γ-test method based on Γ-distribution is put forward for inhomogeneity tests of temperature sequences by researching the mean monthly temperature sequences of 20 stations in Anhui Province from 1961 to 2005. By taking the data of Hefei as an example, the concrete process of testing and analysis is given. The result of Γ-test is compared with the SNHT which is an ordinary method based on normal distribution. The results reveal that this method is superior to SNHT sequences in Scope and accuracy, and it can test most temperature sequences. And it is shown that the data quality of the mean monthly temperature sequences of most stations in Anhui Province is basically reliable, and station moving is the main reason that causes the inhomogeneity of temperature sequences.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2008年第9期1521-1524,共4页
Journal of Hefei University of Technology：Natural Science
基金
中国气象局气候变化专项资助项目(CCSF2007-20)
关键词
变点
Γ检验法
非均一性检验
气温序列
change point
Γ-test method
inhomogeneity test
temperature sequence
