随机误差

即使尽了最大努力，观测永远不会是完美的，因此所有的观测都存在误差和不确定性。不确定性的来源包括：

operator errors (insufficient vigilance,blunders)
population sampling errors (poor sampling design)
instrument errors (glitches or bugs)
calibration errors (zero and span)
instrument limitations (limited resolution or an inappropriate application) 
measurement conditions that are in conflict with the underlying theory.

确定误差的来源，定量误差的性质和量级非常重要，因为：

1. The largest sources of error can be targeted for efforts at error reduction.
2. The uncertainties can be taken into consideration during data analysis and interpretation.The size of the uncertainty may influence how we perceive the data,or the questions to which the data are applied ,as larger uncertainties (or in other words, limited information content) reduce the usefulness of the data.

引自《Eddy Covariance:A Practical Guide to Measurement and Data Analysis》