3.
Interpretation of medium wavelengths
We begin processing of intermediate wavelengths of gravity with three negative anomalies outlined
with dashed lines in Fig. 2 (left). For each area, we reduced anomalies for a model of the regional field
by applying again the upward and downward continuation procedures. The function is a solution of
the Dirichlet’s problem. It satisfies the 2D Laplace equation in the area of investigation and have the
same values at the boundary of the area as the observed field. After subtracting the model of the
regional field, the residuals equal zero at the boundary. Therefore, the upward continuation along the
study area only is possible. According to properties of harmonic functions, it is the smoothest possible
function with known values at the boundary. Besides, this function has no extreme values within the
area, so we do not create artificial anomalies. If we substitute the given data by the regional field
model for each of the three areas, the mentioned negative anomalies are removed. The resulting field
is further inverted for a 3D topography of density interface.
The residual anomalies were approximated by fields of several 3D line segments. The effect of a line
segment was evaluated according to the formula given by Prutkin et al. (2011). To calculate the effect
of a 3D line segment in one observation point, we need only about 2 times more computational time
than that required for a point source. A 3D line segment approximates perfectly an effect of an object
stretched in the some direction. For all three areas, two segments were found to be sufficient to
approximate closely the residual anomalies. Each 3D line segment was described by 7 parameters: 3
coordinates of both ends and the line density. Therefore, we had to determine 14 unknown
parameters using several thousand observations. This procedure was obviously very stable. Further,
we transformed a chosen set of line segments into a restricted object. The approximation based on
line segments provided quite reliable estimates for its mass and a position of its centre of mass.
Fig. 2. Medium wavelengths and discrimination of their sources. Left: medium wavelengths, dashed lines
outline three negative anomalies inside the area. Right: comparison of upper boundaries of Buntsandstein
(Interface 1) and Zechstein (Interface 2) obtained by inversion and according to data from the boreholes (solid
lines with stars and dots, correspondingly).
Two steps of our interpretation methodology (i.e., the subtraction of the regional field model and the
approximation of residuals with the 3D line segments) are demonstrated on the northern negative
gravity anomaly in Fig. 2 (left). The zoomed anomaly, the respective model of the regional field, the
residuals after its subtraction and the results of approximation of the residual field with the field of
two 3D line segments are shown in Fig. 3. The RMS of differences between the residual field and the
effect of line segments is 0.41 mGal.
Next, we inverted the observed gravity (without the negative anomalies) for a 3D topography of
density interface and transformed the sets of line segments into the 3D restricted objects over all
three local areas. In both cases, we applied our inversion algorithms. Inverse problems were reduced
to a nonlinear integral equation of the 1
st
kind for both, the restricted object and the contact surface.
For the restricted body, we assumed its star convexity relative to some interior point (for instance,
relative to its centre of mass). In this way, it was then possible to introduce a spherical coordinate
system (
Dostları ilə paylaş: |