在使用温度脉动仪测量温度结构常数时,平均时间长度的选择会影响其测量结果.通过实际测量数据的分析和讨论,确定平均时间应该为10s左右,以得到真实可靠的结构常数.由于温度脉动仪在测量时会受到各种因素的影响,为了筛选掉不可靠的测量结果,提出了一种用温度脉动原始数据来对测量结果进行筛选的方法.该方法首先排除了实验纪录中的错误测量数据,其次对于异常的实验数据,如某一层结构常数数据的异常偏大或偏小,需要根据双点温差原始数据的频谱分析来确认数据是否正常,以进一步排除异常的测量结果,尽可能保证用以统计分析的数据真实可靠.

To measure the atmospheric optical turbulence , the micro-temperature sensor is often used. But statistic average time can affect the measuring results of temperature structure constants. From analyzing the source data the average time can be set as 5～15 seconds. In the measuring period of micro-temperature , the results will be erroneous when the environmental influence exists or the instrument state is not proper. Most of these situations can be found out from the source data analysis. One method of data differentiation was presented and discussed. First , error data was removed according to the experiment records. Second , if some data was abnormally large or small , the spectrum of source data of temperature difference was used to determine the facticity of data. These two steps could ensure the reliability and authenticity of the data.