This figure is usually quoted in terms dBm, a logarithmic expression for power, referenced to 1 milli-watt.
dBm = 10 log (power / 1 mW)
The response of the instrument to plane waves of all polarisations is usually quoted as being plus or minus 1dB.
To put this in context, consider the following example.
if we elect to calibrate our instrument at say 5mW/cm2, then we can expect that at this received power level, the instrument could present results falling within a given range centred upon 5mW. See below:
5mW = 10 x log(5.0) = 6.9897 dB
Plus 1dB = 7.9897 dB
Converting back to mW:
7.9897 dB = antilog (7.9897 / 10) = 6.295
The upper limit is 6.296 mW/cm2
5mW = 10 x log(5.0) = 6.9897 dB
Minus 1dB = 5.9897 dB
Converting back to mW:
5.9897 dB = antilog (5.9897 / 10) = 3.972
The lower limit is 3.972 mW/cm2
Thus when the instrument is receiving a power level of 5mW the specification would allow the instrument to indicate anything within a range of approximately 1mW either side of 5mW no matter what polarisation the leakage wave presented.
In terms of milli-watts, that is in effect a variation of 20%.
It may be thought that figures produced by instruments that fail to meet the 1dB specification are meaningless since if the instrument fails to pick up in particular polarisations, a potential leakage may be overlooked or missed completely.
A demonstration of the effects of polarisation may be seen by connecting to the internet and entering the following URL:
http://techtv.mit.edu/videos/752-mit-physics-demo—-microwave-polarization
Here, a microwave horn on the left of the screen is transmitting a polarized signal incorporating a sine wave (top trace on the oscilloscope). The signal received by the horn on the right of the screen is shown on the bottom trace of the oscilloscope. Note as the receiver horn is rotated through 90o the signal is no longer detected and the received sine wave disappears despite the fact that the wave is still being transmitted.
Thus, if an instrument used to measure microwave power flux density exhibit the same characteristics, i.e. an ability to receive in only one polarization, there is a strong possibility that the leak from a microwave oven could remain undetected.
The principle is demonstrated well in the URL below. Here, sufficient energy to light a bulp is available only as long as the polarizations of the transmitter and receiver dipoles are aligned.
http://techtv.mit.edu/videos/796-mit-physics-demo—-dipole-antenna