The performance of today's magnetometers is defined by measurement
parameters, including sensitivity, resolution, absolute accuracy,
etc. A list of definitions is provided below to assist you in
understanding how to optimize your surveys, evaluate manufacturer
specifications and clarify often mis-used or interchanged definitions,
such as sensitivity
and resolution. To proceed, choose from any of the following:
Absolute
accuracy
Dead
zone
Gradient
tolerance
Heading error
Reading intervals
(Number of readings per second)
Resolution
Sensitivity
Temperature
range
Tracking
speed
Absolute
accuracy defines maximum deviation from the true value of
the measured magnetic field. Since nobody really knows the true
value of the field, absolute accuracy is determined by considering
factors involved in determining the field value and their accuracy,
like gyromagnetic constant, maximum offset of the time base
frequency etc.
Dead
zone is a set of orientations of the sensor that do not
produce measurements. Optically pumped magnetometers cannot
be oriented either 0 or 90 degrees relative to the magnetic
field direction. Zones 0 0 + 5-10 0 and 90 0 + 5-10 0 are dead
zones for those magnetometers. Some other types of magnetometers,
including Overhauser instruments, operate with omni-directional
(isotropic) sensors.
Gradient
tolerance defines maximum gradient at which the magnetometer
produces meaningful reading, not necessarily with the declared
sensitivity. Gradient tolerance rather defines the limits of
operation of the magnetometer.
Heading
error is a maximum deviation of the measurement in function
of sensor orientation. Sources of heading error may be contamination
of the sensor by magnetic inclusions. In optically pumped magnetometers,
heading errors are related to fundamental physical principles.
More information can be found in "A
Brief Review of Quantum Magnetometers".
Reading
intervals or number of readings per second define
the speed of operation. Sensitivity / accuracy should be defined
at each of intervals as there is an increase in noise that is
not easy to predict mathematically (i.e. does not follow the
general rule that noise is proportional to the square root of
the speed of readings).
Resolution
is a minimum step of the counter used to measure precession
frequency and its conversion into magnetic field. It is generally
substantially higher (an order of magnitude) than the sensitivity
to avoid a contribution of the counter to the overall noise
of the system.
Sensitivity
is a statistical value indicating relative uncertainty of repetitive
readings of the same magnetic field intensity. It is defined
as r.m.s. (root - mean - square) value per square root of a
unit of bandwidth (Hz1/2). For example, a sensitivity of 1 pT
/ Hz1/2 means 1 pT r.m.s. (about 3 - 4 pT peak-to-peak depending
on the character of the noise) will be a scatter of readings
about any "etalon" (fixed value) of the applied magnetic
field per 1 Hz of measurement bandwidth.
For wider
bandwidths the noise is supposed to increase with the square
root of the bandwidth i.e. it should double for 4 Hz of bandwidth
etc. This is not really true for most total field measurements.
It would be more correct to state the noise at a specific number
of readings per second and per associated bandwidth. Sensitivity
only defines the scatter; it does not say anything about systematic
error if measurement or "offset" from the true value
of the magnetic field.
Temperature
range defines the range of temperatures at which the magnetometer
operates with its full specs (sensitivity / accuracy).
Tracking
speed is usually of importance in airborne magnetometry.
It defines how many nT / sec change in the field will be followed
by the magnetometer coherently. The same as gradient tolerance,
tracking speed defines the limits of operation, i.e. in those
conditions maximum sensitivity and / or accuracy is not required.
Comments
or Questions?
We would
appreciate hearing any feedback you may have so that we can
make improvements / additions in future. To send a short email,
click
here.
