8.9 PRACTICAL MEASUREMENT CONSIDERATIONS
A technically suitable method can still produce an unreliable result if the measurement system is incorrectly configured, the source is sampled at an unrepresentative time or position, or important uncertainty contributions are omitted. Practical quality control must therefore extend from preparation through field work, data reduction, and reporting.
The following considerations concern how measurements are performed and documented. Interpretation of source duty cycle, beam steering, traffic loading, and the relationship between a measured value and exposure is addressed separately in Section 8.11.
8.9.1 Instrument Suitability, Calibration, And Field Checks
The complete measurement system should have a current, traceable calibration applicable to the sensor or antenna, instrument, cables, accessories, frequency range, field-strength range, and configuration being used. Calibration at one frequency or level does not establish performance throughout the entire claimed range. Calibration certificates and manufacturer data should be reviewed for correction factors, linearity, isotropy, dynamic range, environmental limits, and stated uncertainty.
Before use, the system should be inspected for damage and correct identification, and its battery status, zero or background indication, self-test, date and time, probe selection, units, detector mode, and overload indication should be checked. Where practicable, response should be confirmed with a known source, field check, or comparison instrument. Checks should be repeated after the survey so that drift, damage, or a failed probe is not discovered only after the results have been reported.
Expected levels should lie comfortably above the noise floor and below overload or saturation. An unexpectedly low reading can indicate insufficient sensitivity or a frequency outside the calibrated band; an apparently stable high reading can indicate detector overload or an out-of-band response. Range changes and overload events should be recorded and the affected measurement repeated.