气象科技论文中常用差值辨析及建议

Discrimination and Suggestions of Common Difference Usage in Meteorological Science and Technology Thesis

为了准确表达科研成果,在气象科技论文写作中须了解各种差值的内涵和区别。常用差值包括绝对误差、相对误差、均方根误差、标准差、距平和方差等。绝对误差并非误差的绝对值,是与真值的差值,其单位与真值相同;相对误差是相对于真值的差值,应以百分数形式出现;均方根误差也是相对于真值的差值,其单位与真值相同;标准差是相对于平均值的差值,其单位与真值相同。通常情况下,这些统计量不能混用,要注意辨析,以准确传意,保证科技论文写作的科学性和逻辑性。

In order to express the scientific research achievements accurately in meteorological thesis, it is necessary to distinguish between the various kinds of error and differences, including absolute error, relative error, root mean square error, standard deviation, departure, variance, etc. Absolute error is not the absolute value of error, but the difference from the true value, whose unit is same as the true value. Relative error is the difference relative to the true value, in the form of percentage. Root mean square error is also the difference relative to the true value, with the same unit of the true value. Standard deviation is difference relative to average value, with the same unit of the true value. Normally, these statistics can not be mixed up, we should pay more attention to discriminating and accurately expressing to ensure the science and technology thesis writing is scientific and logical.