Page 15 - Marine Magnetometer Processing

5. Instruments, Signals and Noise Different types of magnetometer will give different results and this needs to be considered when processing data from each type. Three types of magnetometer are commonly available for marine magnetic surveys; proton, Overhauser and caesium. Each of the three use a different method for measuring the magnetic field and the three methods give results with different properties, but essentially they should all record the same value for magnetic field strength if placed in the same magnetic field. Also, there are differences between instruments of the same type made by different manufacturers and differences between instruments made by the same manufacturer so it is important to understand the strengths and limitations of each one when analysing data from them. The main factors we need to consider are update rate and instrument noise. Update Rate and Sample Interval The update or sample rate of an instrument is the speed at which it can make measurements of the magnetic field: • Proton magnetometers need time to prime their measurement sensor, known as polarizing time, and this can be in the order of one or two seconds. Added to this is a delay in making the measurement so measurements can sometimes only be given once every 2 or 3 seconds. Shorter polarizing times can be used but this reduces the sensitivity of the instrument making it less able to detect small anomalies. • Overhauser magnetometers use a faster method of making measurements and can produce up to four measurements per second (Marine Magnetics SeaSPY). Slower update intervals will increase the sensitivity of the instrument (see below). • Caesium magnetometers can measure up to 40 samples per second (Geometrics G882) but again the quality is improved at lower update intervals. Figure 10 shows the effect of changing the sample rate over the same magnetic anomaly. Graph A shows an anomaly measured using a magnetometer updating at 10Hz or 10 samples per second, we have enough samples or measurements and the shape of the anomaly is clearly represented. Graph B shows the same anomaly measured at 4Hz or 4 samples per second, here the shape of the anomaly is also quite detailed. In Graph C the sample interval has been dropped to one measurement per second and the anomaly is starting to lose its shape as only 6 measurements have been made across it. In the last graph D the samples have been further reduced to once every three seconds as produced by a proton magnetometer, here the anomaly has been reduced to a single value above the background field. This single measurement is indistinguishable from a noise spike and would be rejected as being caused by noise when processing the data. The physical size of the object and the area covered by the magnetic anomaly it creates determines if a particular type of magnetometer will ‘see’ the anomaly. If the anomaly around an object is small then you would need an instrument with a high sample rate to be able to make enough measurements across the anomaly to detect it. For very big anomalies even proton magnetometers can detect them as enough measurements could be made to show that it is a real anomaly rather than just a noise spike. Marine Magnetometer Processing \[14\] © 3H Consulting Ltd