QUANTITATIVE TEXTURAL MEASURES CALCULATION FROM GRIDDED GRAVITY AND MAGNETIC ANOMALY DATA

A textural measure is defined as the distribution of a measure of the spatial variation in amplitudes of a potential field. Measure distributions are determined by computing the measure within a window moving over the data grid. Textural measures computed from gridded magnetic or gravity anomaly data include the distribution of anomaly amplitudes, the number of extrema per unit area, the elongation ratio, the anomaly trend distribution and its variance, the surface area per unit area, the lacunarity, and the multifractal spectrum. The anomaly amplitude distribution reflects the relative contributions of sources to the field. Due to superposition, it can be shifted along the anomaly axis for comparison with distributions from other areas. The number of extrema measures the “noisiness” of the data, and the surface area measures relative amplitudes. Elongation ratio is the ratio of ridges and troughs to all extrema and is determined from local curvature analysis. Trend distribution and its variance are unbiased measures of trends in the field computed from the horizontal gradient. Lacunarity describes the distribution of consistent patches of varying sizes in the data. The multifractal spectrum describes the distribution of scaling exponents (fractional dimensions) in the data. All of these measures are useful in characterizing a lithostratigraphic terrane. To apply these measures practically, subgrids of areas of consistent texture are defined on the basis of the geologic and potential field maps and extracted from the anomaly grid. The textural measures are computed for each subgrid area and the distributions for the various measures compared. In addition to the frequency distribution, a possibility function transformation is computed to produce membership functions of the various measures normalized in the same way to facilitate unbiased comparison of the various textural measures. These can then be used in logical combinations of the various textural measures for decision making. Because “and” and “or” functions are additive for possibility, rather than multiplicative for the frequency, decisions based on possibility are more robust. A regional scale example from the Arctic Ocean aeromagnetic field illustrates the textural measures and their ability to discriminate terranes.