Looking back at our survey using Sensys FGM650 gradiometers has demonstrated just how much our procedures have improved and how clients have benefited. We have gained experience over 6 years of continuous use with Sensys cart-based arrays, initially we thought the system would only be required for larger jobs but with significant improvement in data quality, and the excellent reliability of the system, we were soon using the FGM650s for all survey work.
At the outset we wanted to provide a significant improvement to the resolution of our data so all surveys are carried out at 0.5 traverse spacing with readings at between 0.08m - 0.15m along the traverse. Standard practice is generally 1m traverses with readings at 0.25m. The difference between our old system with the lower resolution data collection and the Sensys system was remarkable. In practice, this gives added confidence to the abstraction and interpretation of anomalies as the morphology of archaeological remains is more accurately defined. The much greater data density allows for more representative processing where data are affected by modern ferrous materials.
The increased resolution also provides more evidence of pit-like features on some sites as well as the location of pits of smaller diameter; there is also far less likelihood of identifying discrete positive anomalies as pits where they are actually small ferrous objects.
We have been fortunate enough to have undertaken a number of resurveys where the initial work had been carried out at the lower 1m x 0.25m resolution and perhaps more critically, surveys had been paced using time-based instruments. The Sensys systems using RTK GNSS positioning always located more archaeological features and often these were either not at all visible in the original work or were hidden in 'noise' due to poor positioning when pacing traverses. Where poor ground conditions were encountered e.g. long vegetation or mud, the difference between the datasets was sometimes incredible. The implications of this are really quite important when it comes to providing a useful archaeological assessment of a site.
Survey procedures are very different to our previous cart-based system which we developed in-house using gradiometers that frequently required balancing. The vibration and stresses to the sensors when mounted to the cart require a very high standard of mechanical stability and strength to ensure the precise vertical alignment of the fluxgates at either end of the gradiometer tube. Many UK developers pursued the use of adjustable fluxgates (mechanically, electronically or both) and although for hand-held systems this seems to work quite well, on carts it's a different matter and they were unable to cope adequately with vibration in our own trials. With the Sensys FGM650, the vertical alignment is fixed using a tension band that provides both the mechanical stability required to minimise heading offsets and provides the strength to cope with cart-based operations. Balancing is effectively achieved during processing which removes any offset value from the data; experience has shown that it is virtually impossible to adequately balance a gradiometer onsite and any electronic or mechanical method of balancing doesn't make up for slight misalignment in the mounting of sensors or through temperature variation - maintaining the optimum vertical alignment of the fluxgates is the most important factor when measuring the vertical magnetic gradient of an archaeological feature. This begs the question as to why onsite mechanical and electronic adjustment was often seen as the best solution for balancing gradiometers? Perhaps the argument for this is historical in that we haven't had the computer processing power that could cope with large area surveys for more than a couple of decades and with single sensor operation the norm until the early 2000s survey costs were prohibitively high so magnetometry was often targeted either after magnetic susceptibility reconnaissance or magnetic scanning. The latter was usually carried out with a fluxgate gradiometer that would require a steady zero value to allow a visual assessment by the operator as to whether anomalies were present or not - who knows what would have been missed using this method! In addition, data were often displayed as X Y plots due to the high cost of graphic printers, the lack of processing power and software to create greyscale images, and perhaps also the ready availability of X Y plotters, so a balanced zero value was desirable to allow the production of meaningful X Y traces from raw data.