Any measurement is meaningless without an uncertainty or tolerance associated with it. All measured numbers contain errors, and as such have an associated uncertainty. Disturbing as it is to an engineer, it is actually not possible to measure anything exactly.
One definition of uncertainty is “A parameter, associated with the result of a measurement that characterizes the dispersion of the values that could be reasonably attributed to the measurand” (the “true” value). The concept of the measurand (what you are trying to assess) is important. A practical definition of uncertainty is “Quantifying the quality of the assessment”. Uncertainty is useful because it makes you think about what could be affecting your result. Uncertainty determination is increasingly being demanded in reports by legislation and by standards.
Uncertainty is a statistical concept, in itself not particularly difficult to understand. However when we are talking about the uncertainty of an environmental noise measurement, things do become tricky since it is difficult to apply statistics to the many real life variables that we are confronted with when we are out on a measurement job.