Polar and azimuthal geometrical divergences in two-dimensional media with a blockwise constant gradient (Q2729331)

The article is devoted to the questions of calculating the geometrical divergence in two-di\-men\-sion\-al media with a blockwise constant gradient which occur in mathematical seismics.NEWLINENEWLINE The main objective of the article is to provide explicit expressions for the geometrical divergence \(l_{\parallel}\) in the plane of a ray and the geometrical divergence \(l_{\perp}\) in the direction perpendicular to the plane of a ray.NEWLINENEWLINE The results obtained read as follows:NEWLINENEWLINE In a medium with a two-dimensional piecewise smooth velocity for each regular \(L\) lying in the transverse plane, the following formula is valid NEWLINE\[NEWLINE \frac{(\dot l_{\perp})_s(M)}{V(M)} = \text{const},\quad M\in L. NEWLINE\]NEWLINE The azimuthal geometrical divergence \(l_{\perp}\) along any ray \(L\) is given by the formula NEWLINE\[NEWLINE l_{\perp}(M_n)\equiv l_{\perp}(M_0,M_n) = \frac{1}{V(M_0)}\int_{M_0}^{M_n} V(M)\,ds_M,\quad M_0, M_n \in L. NEWLINE\]NEWLINE In a two-dimensional block medium with a constant direction \(\vec q_i\) of the velocity gradient \(V_i\) in each block \(B_i\) the following holds: NEWLINE\[NEWLINE \begin{multlined} l_{\perp}(M_n)\equiv l_{\perp}(M_0,M_n) = \frac{1}{V(M_0)}\bigl[V(M_0)l(M_0,M_1) + V^+(M_1)l(M_1,M_2)+ \\ \dots + V^ + (M_{n-1})l(M_{n-1},M_n)\bigr]. \end{multlined} NEWLINE\]NEWLINE Here \(V^+(M_i)\), \(i = 1,\dots, n-1\), are the limiting values of the velocity \(V\) at the points \(M_i\) from the refracted (reflected) part of the ray \(L\).NEWLINENEWLINE The quantity \(l_{\perp}(M_n)\) is calculated by the formula: NEWLINE\[NEWLINE l_{\perp}(M_n) = \frac{1}{V(M_0)}\sum_{i=0}^{n-1}V^+(M_i)\frac{| z_{i+1} - z_i|} {\sin \psi^+ (M_i)}. NEWLINE\]NEWLINE Here \(| z_{i+1} - z_i|\) is the shift of the \(L\) section along the \(Oz\) axis between the points \(M_i\) and \(M_{i+1}\), \(\psi^+(M_i)\) is the angle of the \(Ox\) axis with the tangential vector \(M_i\) to the part of the ray \(L\) going from the point \(M_i\).NEWLINENEWLINE The authors also present an algorithm for calculating the polar geometrical divergence \(l_{\parallel}\) and note that the formulas are easily used for programming.