GNGTS 2017 - 36° Convegno Nazionale

726 GNGTS 2017 S essione 3.3 Methods. In a typical high–resolution magnetic survey, total field magnetic data, T ( x , y ), are usually acquired along a set of survey lines. In the case of moderately disturbed days ( K p = 4) or when the data acquisition requires several hours, it is good practice to correct the data for the daily variations of the geomagnetic field through a levelling procedure. In this instance, it is possible to start with the rapid acquisition along a transverse tie line T 0 that crosses the entire survey area. These data can be considered instantaneous readings at time t = 0, because T 0 is generally travelled in only 1–2 min in the case of archaeological surveys. Then, the survey is performed normally following the survey lines. This method allows to build a diurnal drift function starting from the crossover errors ε i ( t ) = T i ( t ) – T i (0). In most cases, the diurnal drift curve can be obtained fitting a cubic polynomial. After the standard pre–processing step, this curve is then subtracted from the raw data to remove the diurnal variations. The first processing step after despiking, drop–out removal, and levelling should be the calculation of total field values at regular grid locations through a gridding algorithm. Although the general method to obtain magnetic anomalies Δ T ( x , y ) from total field data T ( x , y ) is simply that of subtracting the reference field intensity F ( x , y ) at the same location, so that Δ T ( x , y ) = T ( x , y ) – F ( x , y ), this approach does not generally provide anomalies that are representative of archaeological features. In fact, in this instance the magnetic anomalies will be the expression of an anomalous field Δ F ( x , y ) that includes two sources. A major source, which is not relevant in archaeological studies, is associated with crustal magnetization and has magnitudes of the order of tens to hundreds nT. The signal associated with archaeological objects is generally much smaller, with rms magnitudes not exceeding few tens nT. Consequently, in this instance the procedure to calculate magnetic anomalies from total field data is slightly more complicate, because it is often necessary to isolate a very small–amplitude signal from the observed data. In our approach, magnetic anomalies are calculated subtracting an N degree trend surface from the total field grid value T ( x , y ): (1) Eq. (1) can be justified noting that the Earth’s magnetic field is harmonic in the region outside the Earth’s surface, thereby it has continuous derivatives. Consequently, in any sufficiently small survey area it can be represented by a Taylor’s polynomial series with constant coefficients. Calculation of model anomalies. The computer program ArchaeoMag is designed to operate on UTM georeferenced maps of archaeological anomalies, although in can be also used in local (survey) coordinates. The program assumes that the anomalies have been determined through the correct application of Eq. (1) or a similar method of total field data reduction. In other words, it assumes that the magnetic anomaly amplitudes reflect the true magnetization of the buried archaeological features. In addition to specifying input grids, the user selects a color scale for the representation of the magnetic anomalies and some ambient parameters, which include sensor height, the geomagnetic field parameters ( F , D 0 , I 0 ), and the soil volume susceptibility χ 0 in SI units. Reference field declination, D 0 , and inclination, I 0 , are used to calculate model anomalies starting from anomalous field vectors, while the field intensity, F , is used with the soil susceptibility and the susceptibility of the buried objects to determine the induced component of magnetization M I . This approach clearly requires a preliminary soil sampling and analysis through a magnetic susceptibility meter. Finally, the survey area parameters (corner coordinates and map resolution) are calculated automatically by the program after the specification of an input magnetic anomaly grid. ArchaeoMag allows to define four classes of shapes, corresponding to common archaeological features: 1. spheres (magnetic dipoles), 2. rectangular prisms, 3. generic vertical prisms, and 4. stairways. For any object, the program allows to specify the minimum and maximum burial depths, the magnetic susceptibility, χ, a cutoff distance beyond which the program does not calculate anomalies (for computing time optimization), and a remnant magnetization vector