Calculation of the seismic first-break time field and its ray path distribution using a minimum traveltime tree algorithm

Size: px
Start display at page:

Download "Calculation of the seismic first-break time field and its ray path distribution using a minimum traveltime tree algorithm"

Transcription

1 Geophys. J. Inr. (993) 4, 5936 Calculation of the seismic firstbreak time field and its ray path distribution using a minimum traveltime tree algorithm Shunhua Cao and Stewart Greenhalgh School of Earth Sciences, Flinders Universily of Soulh Australia, GPO Box 2, Adelaide, S.A. 5, Australia Accepted 993 February. Received 993 February ; in original form 992 October 2 SUMMARY A grid search method, analogous to Huygens Principle, is employed to find the firstbreak seismic traveltime field in a seismic model. By defining a set of directions in space, a dynamic directed graph (digraph) containing the firstbreak time information can be constructed during the minimum traveltime treesearching process. The dynamic digraph has far fewer edges than the global graph suggested by previous researchers. An efficient sorting algorithm (heapsort) is adapted to the traveltime data structure by employing an indirect heap strategy. The computational speed can thus be improved several times over other approaches. The minimum traveltime tree algorithm is an efficient and flexible method to simultaneously calculate the firstbreak time field and the corresponding ray paths. It produces a robust and global result in comparison with traditional ray tracing methods. Later seismic phases can also be handled by imposing constraints on the ray paths. A subgrid technique is better than the normal grid technique in terms of accuracy and efficiency. The construction of the digraph uses the local directional information of raypaths and therefore local anisotropic information can be naturely incorporated. The firstbreak time field and its ray paths in an anisotropic model can be calculated as easily as those for an isotropic model. Key words: minimum traveltime, rays, seismic model, tree algorithm. INTRODUCTION Both traveltimes and ray paths are vital information needed for many seismic inverse techniques. Traditionally they are obtained by either ray shooting or ray bending. In shooting, rays are traced from a source at a given initial condition by solving the general ray equations. Many trials have to be made before a ray can reach a preassigned location with an acceptable accuracy. In bending, Fermat s Principle is used in an attempt to find a ray path by minimizing the traveltime from a source to a receiver. It is well known that both methods have severe drawbacks such as slow convergence and local minimum traveltime paths. Recently, two different methods were developed to find the minimum traveltime field for a given source location. The first method is due to Vidale (988, 99) and was improved by Qin et al. (992) and Cao & Greenhalgh (992). This method directly solves the eikonal equation along an expanding wavefront. Since a wavefront is a continuous curve in 2D or surface in 3D, minimum traveltimes from a source to all nodes are found with a Downloaded from on February 28 single run. Therefore, it is extremely efficient for seismic techniques such as prestack migration and hypocentre determination, which only require traveltime information at many locations. Despite the difficulty of incorporating anisotropy into finite difference calculations, we are aware that certain research groups (e.g. Stanford University, University of Texas) are using finite difference methods for traveltime calculations in anisotropic media. However, to the best of our knowledge no details on such formulations have been published. An alternative method of calculating firstarrival times and their ray path distribution is the minimum traveltime tree (shortest path) method. The idea of using the minimum traveltime tree algorithm to track the wavefront and ray paths was originally proposed by Nakanishi & Yamaguchi (986) and further developed by Moser (99). It is based on Fermat s minimum traveltime principle for seismic rays and the algorithms developed in network theory for the construction of the shortest path. Moser (99) also showed that the heapsort algorithm is the most efficient sorting method for the minimum traveltime ray tracing. 593

2 594 S. Cao and S. Greenhalgh In this paper, we wish to show that further iniprovement on the calculation of the minimum traveltime field and the corresponding ray path distribution can be achieved with the construction of a dynamic digraph (directed graph) which is built during the ray searching process rather than a fixed graph constructed at the beginning of the searching process as used by Moser (99). The construction of the dynamic digraph is implicitly based on Huygens Principle (which states that every point on a wavefront can be viewed as a source of secondary waves) because each node in the grid is physically treated as a new source point in evolving the waves through the model. The dynamic digraph has to be constructed with the knowledge of local information about the seismic rays. SHORTESTPATH TREE ALGORITHM The minimum spanning tree algorithm can be found in many computer textbooks under the subject of algorithms (e.g. Baase 988; Sedgewick 988). For reasons of background and convenience of discussion, a brief description is necessarily included here. In this algorithm, a graph (G) is defined as a set of vertices (V) connected by edges (V) and represented as (V, E). A weighted graph or a network is defined as a graph with weighted edges (W) and represented (V, E, W). A spanning tree for graph G is a subgraph of G that is a tree and contains all the vertices of G. In a weighted graph the weight of a subgraph is the sum of the weights of the edges in the subgraph. Obviously a minimum spanning tree of a weighted graph is a collection of edges connecting all the vertices such that the sum of the weights of the edges is at least as small as the sum of the weights of any other collection of edges connecting all the vertices. When the starting vertex (source position) is chosen u priori and cumulative weights from the source are used the minimum spanning tree becomes the shortestpath tree for that particular vertex. Figure schematically illustrates the construction of a shortestpath tree. Each vertex is annotated by a capital letter with a circle. Edges are denoted by solid lines connecting vertices and the number associated with each edge is the weight of the edge. The algorithm begins by selecting an arbitrary starting vertex (a childless tree, see Fig. la). At each iteration, a new vertex and edge is chosen to join the tree constructed so far (see Fig. lb). During the course of the algorithm, the vertices may be thought of as divided into three (disjoint) categories as follows: tree vertices: in the tree constructed so far. Fringe vertices: not in the tree, but adjacent to some vertex in the tree. Unseen vertices: all others. Thus the key step in the algorithm is the selection of a vertex from the fringe vertices. The edge connecting a tree vertex to a fringe vertex is called an incident edge to the fringe vertex. For a sparse graph, priority first search is employed and it always chooses a potential incident edge with the minimum weight (Bellman 958) from all edges between a tree vertex and its associated fringe vertices (see Fig. lc). After each iterative loop of the algorithm, there may be new fringe vertices, and the set of edges from which the next selection is made will change. Fig. l(c) suggests that we (a) A weighted graph (b) The tree and fringe after the starting vertex is selected n Fringe vertices (C)After selecting an edge and vertex. BH is not shown because AH is a better choice (has lower weight) to reach H. Figure. One iteration of the shortestpath tree algorithm. (a) The graph is represented by vertices of circled capital letters and these vertices are connected by solid lines (edges) with the weights specified by the numbers. (b) The vertex A is chosen to be the starting point. (c) After selecting an edge and vertex. BH is not shown because AH is a better choice (the weight of AH, 3 versus the weight of A BH, 4). Downloaded from on February 28

3 Traveltime tree algorithm 595 need not consider all edges between tree vertices and fringe vertices. After AB was chosen, BH became a potential incident edge choice, but it is discarded because AH has lower weight and would be a better choice for reaching H. If the sum of weights from A to H via B was lower than the weight of AH, then AH could be discarded. For each fringe vertex, we need keep track of only one edge to it from the tree, i.e. the one with lowest weight. All such edges are called candidate edges. The shortestpath tree algorithm is outlined below (see also Fig. ). Note that texts in the curly brackets are explanations whereas others are pseudo computer codes. The original work on this algorithm was published in the classical paper of Dijkstra (959). Input: G = (V, E, W), a weighted graph. Output: The edges in a shortestpath tree. () {Initialization} Let x (e.g. A ) be an arbitrary vertex. lirrcc = {x}; ETrcc = {nil}; (2) {Main loop (26); x (e.g. B) has just been brought into the tree. Update fringe vertices and candidate edges. Then add one vertex and edge.} while V,,,,! = V do. (3) {Replace some candidate edges. (see Fig. lc)} for each fringe vertex y (e.g. H ) adjacent to x (i.e. B) do if [W(xy)(e.g. ABH) < W (the candidate edge e incident with y )(e.g. AH)] then xy (i.e. ABH) replaces e (i.e. AH) as the candidate edge for y ; {Note this is for the explanation purpose, and as said in the text the above comparison is not true} end if end for (4) {Find new fringe vertices and candidate edges (see Fig. Ic)} for each unseen y (e.g. C) adjacent to x (i.e. B) do y (i.e. C) is now a fringe vertex; xy (i.e. BC) is now a candidate end for; (5) {Ready to choose next edge} if there are no candidates then stop; else (6) {Choose next edge} Find a candidate edge, e (e.g. AH), with minimum cumulative weight; x (e.g. H) =the fringe vertex incident with e (i.e. AH); Add x (H) and e (AH) to the tree; {x (H) and e (AH) are no longer fringe and candidate.} end if end while The algorithm terminates with V,,,, = V, and the edges in ETrCc form a shortestpath tree. The shortestpath tree can be found in the order of [(E + V ) log V] steps. MINIMUM TRAVELTIME AND ITS RAY PATH DISTRIBUTION Nakanishi & Yamaguchi (986) and Noser (99) showed that with the vertex of the seismic source location as a starting vertex, the minimum traveltime tree gives a very Figure 2. A choice of the predefined searching directions in an 8 x 8 neighbourhood area. Directions are numbered, e.g. one representing a vertically upgoing direction. good approximation to the ray path distribution. The sum of weights from the source to each vertex in the minimum traveltime tree closely approximates the first arrival time distribution. A fixed graph is used in both papers, i.e. a global graph is considered regardless of the weight distribution, or local information. In this paper, the local information is efficiently incorporated into the construction of a dynamic directed graph (digraph). Instead of constructing a graph at the initialization stage, we employ a rule to find the edges of the dynamic digraph during the calculation process. The dynamic digraph is unknown until the minimum traveltime tree is complete. We use the word, dynamic, to refer to the fact that the digraph is growing as the construction of the minimum traveltime tree is progressing. For the convenience of discussion, we use a discrete 2D model to illustrate the methodology. 3D models can be treated in an analogous manner. Figure 2 shows a vertex with its neighbourhood vertices and potential edges. The radius of the neighbourhood is determined by the accuracy requirement, as shown by Moser (99). In this paper, an 8 x 8 square is defined as a neighbourhood for its centre. Because the accuracy is largely determined by the largest angle subtended by two neighbouring edges, we only choose 24 segments as potential edges, although more segments are available in the square. For a global graph definition, these 24 potential edges would be the real edges of the starting graph. For our dynamic digraph, the edges of the graph are selected on the basis of local information. To do this, we have to define a set of rules. First of all, edges are now defined by two parameters: weight and direction. Thus, all potential edges in Fig. 2 denote both the traveltime and direction. Directions are numbered, e.g. one representing a vertical direction. Once the direction of a potential edge and Downloaded from on February 28

4 596 S. Cao and S. Greenhalgh its tail location (i.e. starting point) is defined, the weight or traveltime for this potential edge can be calculated based on the local slowness distribution. Secondly, the edges of the dynamic digraph are determined by the local information, i.e. the direction of the incident edge and the weight distribution, which is controlled by the local slowness variation. Fringe vertices can then be selected according to the minimum traveltime principle. Figure 3 shows that for an incident edge along direction 9 linked with vertex A, edges of the growing digraph are selected based on the local information. Here it is easy to make a comparison with the ray shooting method. If only one potential fringe vertex is allowed, the best choice would be to find a vertex along a direction closest to the direction determined by Snell s Law. In doing so. the minimum :9 traveltime ray tracing degenerates into a multishooting process. To avoid this, we need to search more than one direction (see Fig. 3). There are two approaches to proceed with a number of searching directions. The first approach is to define an angular neighbourhood size according to the local slowness variation. The principle of defining the neighbourhood size is to ensure the direction defined by Snell s Law is within the neighbourhood in most cases. However, it is unnecessary to assume that the direction defined by Snell s Law must fall within the neighbourhood in all cases. If a ray could not turn an angle required by Snell s Law at one vertex, it would continue the turning process at the next vertices until Snell s Law is met. For a small slowness variation, a neighbourhood size of three would be appropriate (see Fig. 3a). Our experience suggests that a neighbourhood size of five is adequate for most situations encountered in seismology. Fig. 3(a) shows that for a smooth medium. a vertex in the growing digraph has three edges which are selected in the neighbourhood of the direction of the incident edge of the vertex. Fig. 3(c) shows that for a medium with a large velocity contrast, a vertex in the growing digraph has five edges which are also selected in the neighbourhood of the direction of the incident edge of the vertex. The second appraoch is to select a direction among the predefined directions closest to the direction defined by Snell s Law, and then search directions in the neighbourhood of the selected direction. Such a procedure provides a superior starting direction to the method above. so the neighbourhood size can be reduced. The local slowness variation in Fig. 3(b) is the same as in Fig. 3(c). but the neighbourhood size is reduced to three whereas its central direction has changed to direction 8. which is closest to the direction defined by Snell s law. The first approach has the advantage of simplicity and it generally requires slightly more computational effort. The second approach is more complicated but more efficient. For a medium having moderate velocity contrast, the first approach is preferred. For a blockwise medium, the second approach is preferred. It is obvious that a substantial computational effort can be saved with the definition of a dynamic digraph which has less edges than a fixed graph. With the definition of 24 directions. three edges are used for a smooth medium and it can perform more than seven times faster than the algorithm with the global graph definition. For a seismic problem with a large velocity contrast, five edges are used and it can still perform more than four times faster than the old algorithm. Greater speed improvement can be gained with a large number of predefined directions, in particular in three dimensions. For example, with 2 predefined spatial directions and search directions, there is approximately a fold speed up over Moser s (99) method. However, it should be noted that when more directions are defined, more searching directions are usually required to ensure that, ~.~.~..~.~~. the actual ray path is within or close to the searching area, especially in strongly heterogeneous regions. Figure 3. Schematic illustration of the choice of edges for a dynamic digraph. (a) Three edges are choen for a smooth model; NUMERICAL IMPLEMENTATION AND (b) three edges are chosen for a model with a large velocity contrast TESTING where the central edge is selected on the basis of Snell s law: and (c) five edges are chosen for a model with a large velocity contrast. To implement the minimum traveltime tree algorithm Direction numbers correspond to those specified in Fig. 2. efficiently. we need to choose a data structure that only Downloaded from on February 28

5 Traveltime tree algorithm 597 M N I J L ID 2 n Figure 4. Data structure for minimum traveltime tree algorithm. stores the necessary information and stores it in such a way that the operations required by the algorithm can be done quickly. First of all, we need three arrays to store slowness, traveltime and raypath information at all the vertices. The remaining data structure is needed to perform the minimum traveltime tree algorithm. Fig. 4 shows in a schematic form the data structure needed for our method. In this data structure, the first two arrays are used to perform indexed heapsort. Next, two (three for 3D) arrays are used to translate data positions in the heap into vertex positions. The fifth array is a status array and functions as a direction indicator which points from the parent vertex. The last one, vertex ID, assigns a sequential number to each datum (label) in the heap. Note that with the above data structure, we already assume a local heap of fringe vertices in two (or three) dimensions instead of a global heap of the whole model as adapted by other researchers with the D formulation of Knuth (968). The total size of the newly defined data structure is only a fraction of the size required by the global D formulation for a 2D model, which will be further explained below. There are two approaches to formulate the heap structure. One is a global strategy. In this approach, the whole traveltime array is organized into a heap structure by indices. At each iteration of the loop, a fringe vertex joins the minimum traveltime tree and the size of the heap decreases by one. Therefore, the size of each of the data arrays has to be of the same size as the model. It should be noted that the number of data arrays in the global approach is less than that in the local approach. A significantly improved approach can be derived from the fact that only fringe vertices are involved at each iteration of the loop. Bearing in mind that a heap structure can be easily expanded by inserting some labels (e.g. traveltime data), or shrunk by removing some labels, we can construct a heap data structure with only the fringe vertices. From iteration to iteration, the size of the data structure may increase or decrease depending on the number of fringe vertices. The maximum size of each of the arrays is determined by the maximum number of fringe vertices at any stage during the process of constructing the minimum traveltime tree. It is obvious that the maximum number of fringe vertices is much smaller than the total number of vertices, especially when working in three dimensions. The vertex ID and ray path information can be stored in the same array because the ray path information is available only after the vertex joins the minimum traveltime tree and the vertex ID is needed only when the vertex is a fringe vertex. Thus the latter approach requires considerably less computer core memory than the global approach. The traveltime spent over each edge is needed only once, and when required it is calculated on the basis of local information. Fringe vertices of each vertex already in the minimum traveltime tree in the dynamic digraph are selected based on the value of the status array at the parent vertex and the local slowness distribution. Theoreticaly, the weight or traveltime for an edge to a fringe vertex should be the minimum traveltime between the two vertices and may be obtained analytically with local slowness information. In practice, straight lines are assumed for all edges. Firstly, a ray path variation has a lower order effect on traveltime than a slowness variation. Secondly, in real applications, many more search directions (e.g. of the order of ) are employed and neighbouring edges are so close that possible discrepancies between the real ray path and the straight line are very small; their effects on traveltime are negligible. Figure S shows the wavefront and ray paths in three models. All the models are of the same size, i.e. 4 x 2 nodes with a node spacing of m. Sources are placed at (, ) in the node coordinate. Wavefronts are illustrated by isochrons with a constant time increment of ms. In Fig. 5(a), the model is homogeneous with a velocity of 5 km s.'; the model in Fig. 5(b) has a high slowness (4kms') zone embedded within a lower slowness (high velocity, i.e. 5 km sc) host material whereas the model in Fig. 5(c) has an embedded low slowness zone (highvelocity inclusion, 6 km si) within a lowvelocity (5 km s') host. The wavefronts near the source in Figs 5(a), (b) and (c) are close to being circular (24sided polygons), which gives a good indication of the homogeneity in the source region. In Fig. 5(a) the wavefronts continue to be circular further away from the source whereas their counterparts in Figs S(b) and (c) are distorted by the velocity inhomogeneity. The wavefronts in the greater slowness region in Fig. S(b) are delayed whereas those in the lower slowness region in Fig. 5(c) are advanced. The ray path distributions in the three models are drastically changed by the different slowness distributions. The minimum traveltime ray paths in Fig. 5(a) are (approximately) evenly distributed. The region having Downloaded from on February 28

6 ') 598 S. Cao and S. Greenhalgh 2 (a' 2 (b' 2 (c) 4 4 higher slowness in Fig. S(b) pushes the minimum traveltime ray paths away from itself, whereas the region with a lower slowness in Fig. S(c) attracts the minimum traveltime ray paths into itself. It is important to note that the distributions of the takeoff angles at the source are changed by the slowness variations. Seismic rays with these takeoff angles are the minimum traveltime ray paths. Ray paths corresponding to other takeoff angles may be obtained with a constrained minimum traveltime ray tracing algorithm, which is beyond the scope of this paper and will be dealt with in a separate paper. It is also important to note that the spatial distribution of minimum traveltime ray paths is not uniform due to the slowness variation; in particular there are no ray paths within the greater slowness (low velocity) zone in Fig. S(b). To facilitate the tracking of ray paths passing through a particular region, accessory interfaces have to be introduced (see e.g. Cao 988; Moser 99). In the case of Fig. S(b), if one wishes to find ray paths passing through the lowvelocity (high slowness) region, an artificial vertical interface may be introduced in the middle of the lowvelocity region of the model. Minimum traveltimes to the artificial interface from any other two points can be obtained and combined to give the time distribution along the artificial interface. If the minimum point on the artificial interface is not at the endpoints, it can then be joined with the other two points to form a ray path, which is a local minimum traveltime ray path between the two points. Fig. S(d) shows a ray tracing from A to B via the lowvelocity region of the model where a vertical accessory interface CD is introduced. (a 2 A 4 B 4 Figure 5. Wavefront and ray diagrams in various models of the same size. 4 X 2(Hl nodes with a node spacing of m. The isochronal time increment is ms. Sources are located at (. ) in the node coordinate. (a) A homogeneous reference model with a velocitv of 5 km s I: (h) an embedded lowvelocity ( km s ~ ' ) inhomogeneity: (c) an embedded highvelocity (6 kin \ ~ inhomogeneitv: and (d) conmained rav tracing from A to B via the artificial interface CD. APPLICATIONS The minimum traveltime tree algorithm is an efficient, robust and global method for tracing rays. Its robustness and global property have been demonstrated in the applications of hypocentre determination in strongly heterogeneous earth models (Mow ef a/. 992b) and in ray tomographic inversion (see e.g. Zhou, Sinadinovaki & Greenhalgh 992). In this paper, we wish to show the incorporation of local information into the technique by performing ray tracing in an anisotropic model. The minimum traveltime tree algorithm can easily incorporate the local directional information of the ray path into the traveltime calculation scheme. The slowness for an edge of the dynamic digraph can be obtained by combining the direction of the edge with the local slowness ellipticity. Fig. 6 shows traveltime contours and ray diagrams for three different models. All the three models are of the same size, i.e. 4 x 2 nodes with a node spacing of m. Sources are located at (2,2) in the node coordinate. The traveltime contour interval is ms. The model in Fig. 6(a) is isotropic with a velocity of 5. km s' and serves as a reference. Velocities for models in Figs 6(b) and (c) are elliptically anisotropic. In Fig. 6(b), the long axis of the velocity anisotropy is horizontal with a value of 6.67 km s', whereas its short axis is vertical with the same value as the reference model. In Fig. 6(c), the long axis of the velocity anisotropy is vertical with the same value as the reference model, whereas its short axis is horizontal with a value of 3.33 km s '. Wavefronts in Figs 6(b) and (c) Downloaded from on February 28

7 Traveltime tree algorithm Figure 6. Wavefront and ray diagrams in elliptically anisotropic models of the same size, 4 X 2 nodes with a node spacing of m. The isochronal time increment is ms. Sources arc located at (2,2) in the node coordinate. (a) An isotropic reference model with a velocity of 5 kms I; (b) the long axis of the velocity anisotropy (6.67 km si) is horizontally orientated, whereas the velocity for the short axis is 5 kmsi; and (c) the long axis of the velocity anisotropy (5 km s ) is vertically orientated, whereas the velocity for the short axis is 3.33 km si.. Figure 7. A choice of predefined searching directions in a staggeredgrid scheme in B 9 X 9 neighhourhood arca. accumulated from the source vertex to the other vertices. However, calculated traveltimes could be smaller than the actual traveltimes for a too coarse sampling of a velocity field, which would allow the shortest paths to bridge a lowvelocity zone (Moser rt al. 992a). While the relative error is not related to the model size, the absolute error does vary with the model size. Therefore, it is desirable to subtract from the calculated traveltime the mean error associated with the space and angle discretization and the model size. The mean errors for complicated models are difficult to obtain; an estimate can be made using a corresponding homogeneous model. clearly show the effect of the velocity anisotropy, i.e. wavefronts are elongated horizontally in Figs 6(b) and vertically in Fig. 6(c). ERROR ANALYSIS AND DISCUSSION Errors associated with the traveltimes calculated with the minimum traveltime tree algorithm are dependent upon the grid and angular spacings. They are correct up to the second order in the space and angle discretization, as shown by Moser (99). It should be pointed out that calculated traveltimes will be greater (in a statistical sense) than the actual traveltimes due to the way in which the traveltime is..! Figure 8. A subgrid technique. (a) Only filled circles are timed and open circles are ignored; and (b) potential edges for the dynamic digraph. Downloaded from on February 28

8 6 S. Cao and S. Greenhalgh....,L\. :to) Figure 6. (Confinued.) To reduce the traveltime error, either the grid spacing or the angle spacing or both have to be reduced. This requires more computational effort. In particular, a small angle spacing will require the definition of a large neighbourhood. The large size of the neighbourhood will in turn create a large number of fringe vertices during the construction of the minimum traveltime tree, which then requires more computer time. An improvement can be made by arranging the vertex distribution in a staggeredgrid fashion. Fig. 7 shows the potential edge distribution in a 9 x 9 square neighbourhood of a staggered grid space discretization. In comparison with Fig. 2, many more searching directions are available and the angular spacing is more than halved. When the size of the neighbourhood is of the order of a wavelength or less, a subgrid technique (Nakanishi & Yamaguchi 986) can be used to decrease the number of fringe vertices. Fig. 8(a) shows a 7 x 7 neighbourhood vertex distribution where a staggeredgrid technique is employed to rearrange the vertex distribution. Slowness is assumed constant in each quadrant of the neighbourhood. In the subgrid technique, only peripheral grids (filled circles) in each quadrant of the neighbourhood are considered and inside grids (open circles) are ignored based on the fact that the ray path variation has a lesser effect on traveltime than the slowness variation. Fig. 8(b) shows the potential edge distribution for the neighbourhood shown in Fig. 8(a). The total number of potential edges in the subgrid technique is obviously much smaller than that used for a normal search. With less grid points, the computational speed can be improved, especially when the number of vertices in the neighbourhood is very large. However, this subgrid technique would involve considerably more programming effort because several edges with different lengths are possible for the same spatial direcion. CONCLUSIONS A numerically efficient algorithm for the calculation of the firstbreak seismic traveltime field and its ray path distribution is obtained by using local information during the construction of the dynamic digraph. By discretizing the space in a staggeredgrid manner, better accuracy can be achieved. A subgrid technique can be used to further improve the efficiency of the algorithm when the spatial interval is of the order of a subwavelength or less. The robustness and global nature of the scheme are inherent advantages of the shortestpath tree algorithm. The capability to incorporate local information is illustrated by ray tracing through anisotropic models. The simultaneous solution of traveltimes and ray paths makes this approach very attractive for inclusion in many seismic inversion techniques. ACKNOWLEDGMENT This research was supported by a grant from the Australian Research Council. REFERENCES Baase, S., 988. Computer algorithms, AddisonWesley, Reading, MA. Bellman, R., 958. On a routing problem, Q. appl. Math., 6, 88YO. Cao, S., 988. Rays and isochrons, in PhD midterm essay. Research School of Earth Sciences, Australian National University. Cao, S. & Greenhalgh, S. A,, 992. Finitedifference solution of the eikonal equation using an efficient firstarrival wavefront tracking scheme, Geophysics, submitted. Dijkstra, E. W., 959. A note on two problcms in connexion with graphs, Numer. Math.,, Knuth, D. E., 968. The urt of computer programming, Vol. 3: sorting and searching. AddisonWesley, Reading, MA. Moser, T. J., 99. Shortest path calculation of seismic rays, Geophysics, 56, 596. Moser. T. J., Nolet, G. & Sneider, R., 992a. Ray bending revisited, Bull. seism. Soc. Am., 82, Moser, T. J., van Eck, T. & Nolet, G., 992b. Hypocenter determination in strongly heterogeneous Earth models using the shortest path method, J. geophys. Res., 97, Nakanishi, I. & Yamaguchi, K., 986. A numerical experiment on nonlinear image reconstruction from firstarrival times for twodimensional island are structure, J. Phys. Earth, 34, 952. Qin, F., Luo, Y., Olsen, K. B., Cai. W. & Schuster, G. T., 992. Finitedifference soluton of the eikonal equation along expanding wavefronts, Geophysics, 57, Sedgewick, R., 988. Algorithms, AddisonWesley, Reading, MA. Vidale. J., 988. Finitedifference calculation of travel times, Bull. seism. SOC. Am. 78, Vidale, J., 99. Finitedifference calculation of traveltimes in three dimensions, Geophysics, 55, Zhou, B., Sinadinovski, C. & Greenhalgh, S., 992. Nonlinear inversion traveltime tomography: imaging highcontrast inhomogeneitics, Expl. Geophys. 23, Downloaded from on February 28

Shortest-path calculation of first arrival traveltimes by expanding wavefronts

Shortest-path calculation of first arrival traveltimes by expanding wavefronts Stanford Exploration Project, Report 82, May 11, 2001, pages 1 144 Shortest-path calculation of first arrival traveltimes by expanding wavefronts Hector Urdaneta and Biondo Biondi 1 ABSTRACT A new approach

More information

Fermat s principle and ray tracing in anisotropic layered media

Fermat s principle and ray tracing in anisotropic layered media Ray tracing in anisotropic layers Fermat s principle and ray tracing in anisotropic layered media Joe Wong ABSTRACT I consider the path followed by a seismic signal travelling through velocity models consisting

More information

GEOPHYS 242: Near Surface Geophysical Imaging. Class 4: First-Arrival Traveltime Tomography Mon, April 11, 2011

GEOPHYS 242: Near Surface Geophysical Imaging. Class 4: First-Arrival Traveltime Tomography Mon, April 11, 2011 GEOPHYS 242: Near Surface Geophysical Imaging Class 4: First-Arrival Traveltime Tomography Mon, April 11, 2011 Wavefront tracing methods - speed versus accuracy Inversion algorithms - any magic approach?

More information

A NUMERICAL EXPERIMENT ON NONLINEAR IMAGE RECONSTRUCTION FROM FIRST-ARRIVAL TIMES FOR TWO-DIMENSIONAL ISLAND ARC STRUCTURE

A NUMERICAL EXPERIMENT ON NONLINEAR IMAGE RECONSTRUCTION FROM FIRST-ARRIVAL TIMES FOR TWO-DIMENSIONAL ISLAND ARC STRUCTURE Letter J. Phys. Earth, 34, 195-201, 1986 A NUMERICAL EXPERIMENT ON NONLINEAR IMAGE RECONSTRUCTION FROM FIRST-ARRIVAL TIMES FOR TWO-DIMENSIONAL ISLAND ARC STRUCTURE Ichiro NAKANISHI and Kaoru YAMAGUCHI

More information

Travel Time Tomography using Neural Networks

Travel Time Tomography using Neural Networks Travel Time Tomography using Neural Networks Yoshiya Oda Tokyo Metropolitan University, Japan Tomohisa Ishiyama Tokyo Metropolitan University, Japan Shinya Yokono Tokyo Metropolitan University, Japan SUMMARY:

More information

Vidale, J., 1988, Finite-difference calculation of travel times, BSSA, Vo. 78, No. 6,

Vidale, J., 1988, Finite-difference calculation of travel times, BSSA, Vo. 78, No. 6, Class 4: First-Arrival Traveltime Tomography Mon, Sept 21, 2009 Wavefront tracing methods Inversion algorithms Model regularization Review of tomography case histories, presentation by Dr. Xianhuai Zhu

More information

3-D traveltime computation using Huygens wavefront tracing

3-D traveltime computation using Huygens wavefront tracing GEOPHYSICS, VOL. 66, NO. 3 MAY-JUNE 2001); P. 883 889, 12 FIGS., 1 TABLE. 3-D traveltime computation using Huygens wavefront tracing Paul Sava and Sergey Fomel ABSTRACT Traveltime computation is widely

More information

UNIVERSITY OF CALGARY. Traveltime Tomography in Isotropic and Transversely Isotropic Media. Marco Perez

UNIVERSITY OF CALGARY. Traveltime Tomography in Isotropic and Transversely Isotropic Media. Marco Perez Important Notice This copy may be used only for the purposes of research and private study, and any use of the copy for a purpose other than research or private study may require the authorization of the

More information

(x, y, z) m 2. (x, y, z) ...] T. m 2. m = [m 1. m 3. Φ = r T V 1 r + λ 1. m T Wm. m T L T Lm + λ 2. m T Hm + λ 3. t(x, y, z) = m 1

(x, y, z) m 2. (x, y, z) ...] T. m 2. m = [m 1. m 3. Φ = r T V 1 r + λ 1. m T Wm. m T L T Lm + λ 2. m T Hm + λ 3. t(x, y, z) = m 1 Class 1: Joint Geophysical Inversions Wed, December 1, 29 Invert multiple types of data residuals simultaneously Apply soft mutual constraints: empirical, physical, statistical Deal with data in the same

More information

Prestack Kirchhoff time migration for complex media

Prestack Kirchhoff time migration for complex media Stanford Exploration Project, Report 97, July 8, 998, pages 45 6 Prestack Kirchhoff time migration for complex media Tariq Alkhalifah keywords: time migration, anisotropy ABSTRACT Constructing the seismic

More information

Crosswell Imaging by 2-D Prestack Wavepath Migration

Crosswell Imaging by 2-D Prestack Wavepath Migration Crosswell Imaging by 2-D Prestack Wavepath Migration Hongchuan Sun ABSTRACT Prestack wavepath migration (WM) is applied to 2-D synthetic crosswell data, and the migrated images are compared to those from

More information

96 Alkhalifah & Biondi

96 Alkhalifah & Biondi Stanford Exploration Project, Report 97, July 8, 1998, pages 95 116 The azimuth moveout operator for vertically inhomogeneous media Tariq Alkhalifah and Biondo L. Biondi 1 keywords: AMO, DMO, dip moveout

More information

Parallelization of the seismic ray trace algorithm

Parallelization of the seismic ray trace algorithm Parallelization of the seismic ray trace algorithm Kamil Szostek 1 and Andrzej Le±niak 1 AGH University of Science and Technology, Faculty of Geology, Geophysics and Environment Protection, Department

More information

GEOPHYS 242: Near Surface Geophysical Imaging. Class 8: Joint Geophysical Inversions Wed, April 20, 2011

GEOPHYS 242: Near Surface Geophysical Imaging. Class 8: Joint Geophysical Inversions Wed, April 20, 2011 GEOPHYS 4: Near Surface Geophysical Imaging Class 8: Joint Geophysical Inversions Wed, April, 11 Invert multiple types of data residuals simultaneously Apply soft mutual constraints: empirical, physical,

More information

P063 ADAPTIVE TRAVEL-TIME AND RAY- PARAMETER INVERSION OF DENSELY SAMPLED 2-D SEISMIC DATA

P063 ADAPTIVE TRAVEL-TIME AND RAY- PARAMETER INVERSION OF DENSELY SAMPLED 2-D SEISMIC DATA 1 P063 ADAPTIVE TRAVEL-TIME AND RAY- PARAMETER INVERSION OF DENSELY SAMPLED 2-D SEISMIC DATA I. TRINKS 1, S. SINGH 2, C. CHAPMAN 3, P. BARTON 1, M. BOSCH 4, A. CHERRETT 5 Addresses 1 Bullard Laboratories,

More information

IMAGING USING MULTI-ARRIVALS: GAUSSIAN BEAMS OR MULTI-ARRIVAL KIRCHHOFF?

IMAGING USING MULTI-ARRIVALS: GAUSSIAN BEAMS OR MULTI-ARRIVAL KIRCHHOFF? IMAGING USING MULTI-ARRIVALS: GAUSSIAN BEAMS OR MULTI-ARRIVAL KIRCHHOFF? Summary Samuel H. Gray* Veritas DGC Inc., 715 Fifth Ave. SW, Suite 2200, Calgary, AB Sam_gray@veritasdgc.com Carl Notfors Veritas

More information

GEOPHYS 242: Near Surface Geophysical Imaging. Class 5: Refraction Migration Methods Wed, April 13, 2011

GEOPHYS 242: Near Surface Geophysical Imaging. Class 5: Refraction Migration Methods Wed, April 13, 2011 GEOPHYS 242: Near Surface Geophysical Imaging Class 5: Refraction Migration Methods Wed, April 13, 2011 Migration versus tomography Refraction traveltime and wavefield migration The theory of interferometry

More information

Ray-based tomography with limited picking

Ray-based tomography with limited picking Stanford Exploration Project, Report 110, September 18, 2001, pages 1 113 Short Note Ray-based tomography with limited picking Robert G. Clapp 1 INTRODUCTION In ray-based reflection tomography picking

More information

Efficient Beam Velocity Model Building with Tomography Designed to Accept 3d Residuals Aligning Depth Offset Gathers

Efficient Beam Velocity Model Building with Tomography Designed to Accept 3d Residuals Aligning Depth Offset Gathers Efficient Beam Velocity Model Building with Tomography Designed to Accept 3d Residuals Aligning Depth Offset Gathers J.W.C. Sherwood* (PGS), K. Sherwood (PGS), H. Tieman (PGS), R. Mager (PGS) & C. Zhou

More information

HOUGH TRANSFORM CS 6350 C V

HOUGH TRANSFORM CS 6350 C V HOUGH TRANSFORM CS 6350 C V HOUGH TRANSFORM The problem: Given a set of points in 2-D, find if a sub-set of these points, fall on a LINE. Hough Transform One powerful global method for detecting edges

More information

Separation of diffracted waves in TI media

Separation of diffracted waves in TI media CWP-829 Separation of diffracted waves in TI media Yogesh Arora & Ilya Tsvankin Center for Wave Phenomena, Colorado School of Mines Key words: Diffractions, Kirchhoff, Specularity, Anisotropy ABSTRACT

More information

Multi-azimuth velocity estimation

Multi-azimuth velocity estimation Stanford Exploration Project, Report 84, May 9, 2001, pages 1 87 Multi-azimuth velocity estimation Robert G. Clapp and Biondo Biondi 1 ABSTRACT It is well known that the inverse problem of estimating interval

More information

u = v is the Laplacian and represents the sum of the second order derivatives of the wavefield spatially.

u = v is the Laplacian and represents the sum of the second order derivatives of the wavefield spatially. Seismic Elastic Modelling Satinder Chopra, Arcis Corporation, Calgary, Canada Summary Seismic modeling experiments were performed to understand the seismic responses associated with a variety of subsurface

More information

Reflectivity modeling for stratified anelastic media

Reflectivity modeling for stratified anelastic media Reflectivity modeling for stratified anelastic media Peng Cheng and Gary F. Margrave Reflectivity modeling ABSTRACT The reflectivity method is widely used for the computation of synthetic seismograms for

More information

H003 Deriving 3D Q Models from Surface Seismic Data Using Attenuated Traveltime Tomography

H003 Deriving 3D Q Models from Surface Seismic Data Using Attenuated Traveltime Tomography H003 Deriving 3D Q Models from Surface Seismic Data Using Attenuated Traveltime Tomography M. Cavalca* (Schlumberger - Westerngeco) & R.P. Fletcher (Schlumberger - Westerngeco) SUMMARY Estimation of the

More information

Processing converted-wave data in the tau-p domain: rotation toward the source and moveout correction

Processing converted-wave data in the tau-p domain: rotation toward the source and moveout correction τ-p domain converted-wave processing Processing converted-wave data in the tau-p domain: rotation toward the source and moveout correction Raul Cova and Kris Innanen ABSTRACT The asymmetry of the converted-wave

More information

Angle Gathers for Gaussian Beam Depth Migration

Angle Gathers for Gaussian Beam Depth Migration Angle Gathers for Gaussian Beam Depth Migration Samuel Gray* Veritas DGC Inc, Calgary, Alberta, Canada Sam Gray@veritasdgc.com Abstract Summary Migrated common-image-gathers (CIG s) are of central importance

More information

SEG/New Orleans 2006 Annual Meeting

SEG/New Orleans 2006 Annual Meeting and its implications for the curvelet design Hervé Chauris, Ecole des Mines de Paris Summary This paper is a first attempt towards the migration of seismic data in the curvelet domain for heterogeneous

More information

Finite-difference calculation of direct-arrival traveltimes using the eikonal equation

Finite-difference calculation of direct-arrival traveltimes using the eikonal equation GEOPHYSICS, VOL. 67, NO. 4 (JULY-AUGUST 2002); P. 1270 1274, 9 FIGS. 10.1190/1.1500389 Finite-difference calculation of direct-arrival traveltimes using the eikonal equation Le-Wei Mo and Jerry M. Harris

More information

Angle-gather time migration a

Angle-gather time migration a Angle-gather time migration a a Published in SEP report, 1, 141-15 (1999) Sergey Fomel and Marie Prucha 1 ABSTRACT Angle-gather migration creates seismic images for different reflection angles at the reflector.

More information

The Interface Fresnel Zone revisited

The Interface Fresnel Zone revisited The Interface Fresnel Zone revisited Nathalie Favretto-Cristini, Paul Cristini and Eric de Bazelaire ABSTRACT We determine the part of reflectors which actually affects the reflected wavefield, which is

More information

Interpolation within ray tubes - state of the art

Interpolation within ray tubes - state of the art Interpolation within ray tubes - state of the art Petr Bulant Department of Geophysics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 3, 121 16 Praha 2, Czech Republic, http://sw3d.cz/staff/bulant.htm

More information

Fast marching methods

Fast marching methods 1 Fast marching methods Lecture 3 Alexander & Michael Bronstein tosca.cs.technion.ac.il/book Numerical geometry of non-rigid shapes Stanford University, Winter 2009 Metric discretization 2 Approach I:

More information

Seismic modelling with the reflectivity method

Seismic modelling with the reflectivity method Yongwang Ma, Luiz Loures, and Gary F. Margrave ABSTRACT Seismic modelling with the reflectivity method Numerical seismic modelling is a powerful tool in seismic imaging, interpretation and inversion. Wave

More information

P H Y L A B 1 : G E O M E T R I C O P T I C S

P H Y L A B 1 : G E O M E T R I C O P T I C S P H Y 1 4 3 L A B 1 : G E O M E T R I C O P T I C S Introduction Optics is the study of the way light interacts with other objects. This behavior can be extremely complicated. However, if the objects in

More information

Raypath and traveltime computations Main for 2D transversely isotropic media with Heading dipping symmetry axes Bing Zhou Authors Stewart Greenhalgh

Raypath and traveltime computations Main for 2D transversely isotropic media with Heading dipping symmetry axes Bing Zhou Authors Stewart Greenhalgh Exploration Geophysics (2006) 37, 150 159 Raypath and traveltime computations Main for 2D transversely isotropic media with Heading dipping symmetry axes Bing Zhou Authors Stewart Greenhalgh Key Words:

More information

Applications of the tau-p Inversion Method. Shoba Maraj

Applications of the tau-p Inversion Method. Shoba Maraj Applications of the tau-p Inversion Method Shoba Maraj Final Report Theoretical Seismology Geop 523 Due 6th May 2011 Introduction Generally, velocity increases with depth in the Earth. They are related

More information

=, (1) SEG/New Orleans 2006 Annual Meeting

=, (1) SEG/New Orleans 2006 Annual Meeting U. Albertin* +, P. Sava ++, J. Etgen +, and M. Maharramov + + BP EPTG, Houston, Texas, ++ Colorado School of Mines, Goldin, Colorado Summary A methodology for velocity updating using one-way wavefield

More information

dp i = u ray c2 dt with t the travel time.

dp i = u ray c2 dt with t the travel time. Three dimensional two point ray tracing using paraxial rays in cartessian coordinates Ronit Strahilevitz*, Paradigm Geophysical; Dan Kosloff, Department of Geophysics, Tel Aviv University, Israel and Paradigm

More information

A new method for determination of a wave-ray trace based on tsunami isochrones

A new method for determination of a wave-ray trace based on tsunami isochrones Bull. Nov. Comp. Center, Math. Model. in Geoph., 13 (2010), 93 101 c 2010 NCC Publisher A new method for determination of a wave-ray trace based on tsunami isochrones An.G. Marchuk Abstract. A new method

More information

Advances in radial trace domain coherent noise attenuation

Advances in radial trace domain coherent noise attenuation Advances in radial trace domain coherent noise attenuation ABSTRACT David C. Henley* CREWES, Department of Geology and Geophysics University of Calgary, Calgary, AB henley@crewes.org The radial trace transform,

More information

SEG/New Orleans 2006 Annual Meeting

SEG/New Orleans 2006 Annual Meeting 3-D tomographic updating with automatic volume-based picking Dimitri Bevc*, Moritz Fliedner, Joel VanderKwaak, 3DGeo Development Inc. Summary Whether refining seismic images to evaluate opportunities in

More information

Reverse time migration in midpoint-offset coordinates

Reverse time migration in midpoint-offset coordinates Stanford Exploration Project, Report 111, June 9, 00, pages 19 156 Short Note Reverse time migration in midpoint-offset coordinates Biondo Biondi 1 INTRODUCTION Reverse-time migration (Baysal et al., 198)

More information

Optimised corrections for finite-difference modelling in two dimensions

Optimised corrections for finite-difference modelling in two dimensions Optimized corrections for 2D FD modelling Optimised corrections for finite-difference modelling in two dimensions Peter M. Manning and Gary F. Margrave ABSTRACT Finite-difference two-dimensional correction

More information

A comparison of exact, approximate, and linearized ray tracing methods in transversely isotropic media

A comparison of exact, approximate, and linearized ray tracing methods in transversely isotropic media Ray Tracing in TI media A comparison of exact, approximate, and linearized ray tracing methods in transversely isotropic media Patrick F. Daley, Edward S. Krebes, and Laurence R. Lines ABSTRACT The exact

More information

Comparison of two P-S conversion-point mapping approaches for Vertical Transversely Isotropic (VTI) media

Comparison of two P-S conversion-point mapping approaches for Vertical Transversely Isotropic (VTI) media Raytracing in VTI media Comparison of two P-S conversion-point mapping approaches for Vertical Transversely Isotropic (VTI) media Jianli Yang and Don C. Lawton ABSTRACT Determination of the conversion

More information

Comparison of two P-S conversion-point mapping approaches for Vertical Transversely Isotropic (VTI) media

Comparison of two P-S conversion-point mapping approaches for Vertical Transversely Isotropic (VTI) media Comparison of two P-S conversion-point mapping approaches for Vertical Transversely Isotropic (VTI) media Jianli Yang* and Don C. Lawton CREWES, University of Calgary, 5 UniversityDR.NW, Calgary, AB, TN

More information

Summary. Figure 1: Simplified CRS-based imaging workflow. This paper deals with the boxes highlighted in green.

Summary. Figure 1: Simplified CRS-based imaging workflow. This paper deals with the boxes highlighted in green. Smoothing and automated picking of kinematic wavefield attributes Tilman Klüver and Jürgen Mann, Geophysical Institute, University of Karlsruhe, Germany Copyright 2005, SBGf Sociedade Brasiliera de Geofísica

More information

Coherent partial stacking by offset continuation of 2-D prestack data

Coherent partial stacking by offset continuation of 2-D prestack data Stanford Exploration Project, Report 82, May 11, 2001, pages 1 124 Coherent partial stacking by offset continuation of 2-D prestack data Nizar Chemingui and Biondo Biondi 1 ABSTRACT Previously, we introduced

More information

Adaptive Waveform Inversion: Theory Mike Warner*, Imperial College London, and Lluís Guasch, Sub Salt Solutions Limited

Adaptive Waveform Inversion: Theory Mike Warner*, Imperial College London, and Lluís Guasch, Sub Salt Solutions Limited Adaptive Waveform Inversion: Theory Mike Warner*, Imperial College London, and Lluís Guasch, Sub Salt Solutions Limited Summary We present a new method for performing full-waveform inversion that appears

More information

Chapter 2 Methodology of Seismic Tomography

Chapter 2 Methodology of Seismic Tomography Chapter Methodology of Seismic Tomography Abstract In this chapter we introduce the tomographic methods which are widely used to study three-dimensional (3-D) seismic velocity attenuation and anisotropy

More information

Anisotropic model building with well control Chaoguang Zhou*, Zijian Liu, N. D. Whitmore, and Samuel Brown, PGS

Anisotropic model building with well control Chaoguang Zhou*, Zijian Liu, N. D. Whitmore, and Samuel Brown, PGS Anisotropic model building with well control Chaoguang Zhou*, Zijian Liu, N. D. Whitmore, and Samuel Brown, PGS Summary Anisotropic depth model building using surface seismic data alone is non-unique and

More information

Estimating interval shear-wave splitting from multicomponent virtual shear check shots

Estimating interval shear-wave splitting from multicomponent virtual shear check shots GEOPHYSICS, VOL. 73, NO. 5 SEPTEMBER-OCTOBER 2008 ; P. A39 A43, 5 FIGS. 10.1190/1.2952356 Estimating interval shear-wave splitting from multicomponent virtual shear check shots Andrey Bakulin 1 and Albena

More information

SEG/New Orleans 2006 Annual Meeting

SEG/New Orleans 2006 Annual Meeting Accuracy improvement for super-wide angle one-way waves by wavefront reconstruction Ru-Shan Wu* and Xiaofeng Jia, Modeling and Imaging Laboratory, IGPP, University of California, Santa Cruz Summary To

More information

Successful application of joint reflection/refraction tomographic velocity inversion in a shallow water marine environment.

Successful application of joint reflection/refraction tomographic velocity inversion in a shallow water marine environment. Successful application of joint reflection/refraction tomographic velocity inversion in a shallow water marine environment. Sergey Birdus 1, Dean Criddle 2, Alexey Artyomov 1, Li Li 1, Qingbing Tang 1,

More information

2010 SEG SEG Denver 2010 Annual Meeting

2010 SEG SEG Denver 2010 Annual Meeting Localized anisotropic tomography with checkshot : Gulf of Mexico case study Andrey Bakulin*, Yangjun (Kevin) Liu, Olga Zdraveva, WesternGeco/Schlumberger Summary Borehole information must be used to build

More information

We N Converted-phase Seismic Imaging - Amplitudebalancing Source-independent Imaging Conditions

We N Converted-phase Seismic Imaging - Amplitudebalancing Source-independent Imaging Conditions We N106 02 Converted-phase Seismic Imaging - Amplitudebalancing -independent Imaging Conditions A.H. Shabelansky* (Massachusetts Institute of Technology), A.E. Malcolm (Memorial University of Newfoundland)

More information

TRAVELTIME TOMOGRAPHY (CONT)

TRAVELTIME TOMOGRAPHY (CONT) 30 March 005 MODE UNIQUENESS The forward model TRAVETIME TOMOGRAPHY (CONT di = A ik m k d = Am (1 Data vecto r Sensitivit y matrix Model vector states the linearized relationship between data and model

More information

UNIT 2 GRAPHIC PRIMITIVES

UNIT 2 GRAPHIC PRIMITIVES UNIT 2 GRAPHIC PRIMITIVES Structure Page Nos. 2.1 Introduction 46 2.2 Objectives 46 2.3 Points and Lines 46 2.4 Line Generation Algorithms 48 2.4.1 DDA Algorithm 49 2.4.2 Bresenhams Line Generation Algorithm

More information

mywbut.com Diffraction

mywbut.com Diffraction Diffraction If an opaque obstacle (or aperture) is placed between a source of light and screen, a sufficiently distinct shadow of opaque (or an illuminated aperture) is obtained on the screen.this shows

More information

We B3 12 Full Waveform Inversion for Reservoir Characterization - A Synthetic Study

We B3 12 Full Waveform Inversion for Reservoir Characterization - A Synthetic Study We B3 12 Full Waveform Inversion for Reservoir Characterization - A Synthetic Study E. Zabihi Naeini* (Ikon Science), N. Kamath (Colorado School of Mines), I. Tsvankin (Colorado School of Mines), T. Alkhalifah

More information

specular diffuse reflection.

specular diffuse reflection. Lesson 8 Light and Optics The Nature of Light Properties of Light: Reflection Refraction Interference Diffraction Polarization Dispersion and Prisms Total Internal Reflection Huygens s Principle The Nature

More information

Chapter 3. Sukhwinder Singh

Chapter 3. Sukhwinder Singh Chapter 3 Sukhwinder Singh PIXEL ADDRESSING AND OBJECT GEOMETRY Object descriptions are given in a world reference frame, chosen to suit a particular application, and input world coordinates are ultimately

More information

2D rendering takes a photo of the 2D scene with a virtual camera that selects an axis aligned rectangle from the scene. The photograph is placed into

2D rendering takes a photo of the 2D scene with a virtual camera that selects an axis aligned rectangle from the scene. The photograph is placed into 2D rendering takes a photo of the 2D scene with a virtual camera that selects an axis aligned rectangle from the scene. The photograph is placed into the viewport of the current application window. A pixel

More information

Angle-dependent reflectivity by wavefront synthesis imaging

Angle-dependent reflectivity by wavefront synthesis imaging Stanford Exploration Project, Report 80, May 15, 2001, pages 1 477 Angle-dependent reflectivity by wavefront synthesis imaging Jun Ji 1 ABSTRACT Elsewhere in this report, Ji and Palacharla (1994) show

More information

Data dependent parameterization and covariance calculation for inversion of focusing operators

Data dependent parameterization and covariance calculation for inversion of focusing operators Stanford Exploration Project, Report 11, September 18, 21, pages 1 91 Data dependent parameterization and covariance calculation for inversion of focusing operators Barbara E. Cox 1 ABSTRACT The Common

More information

COMP30019 Graphics and Interaction Scan Converting Polygons and Lines

COMP30019 Graphics and Interaction Scan Converting Polygons and Lines COMP30019 Graphics and Interaction Scan Converting Polygons and Lines Department of Computer Science and Software Engineering The Lecture outline Introduction Scan conversion Scan-line algorithm Edge coherence

More information

Migration Resolution. Chapter Time Migration vs Depth Migration

Migration Resolution. Chapter Time Migration vs Depth Migration Chapter 3 Migration Resolution In this section time migration is defined and its pitfalls and benefits are compared to depth migration. In the far-field approximation the resolution formula for analyzing

More information

22 Elementary Graph Algorithms. There are two standard ways to represent a

22 Elementary Graph Algorithms. There are two standard ways to represent a VI Graph Algorithms Elementary Graph Algorithms Minimum Spanning Trees Single-Source Shortest Paths All-Pairs Shortest Paths 22 Elementary Graph Algorithms There are two standard ways to represent a graph

More information

Migration from a non-flat datum via reverse-time extrapolation

Migration from a non-flat datum via reverse-time extrapolation Stanford Exploration Project, Report 84, May 9, 2001, pages 1 50 Migration from a non-flat datum via reverse-time extrapolation Gopal Palacharla 1 ABSTRACT Land surveys usually have elevation changes,

More information

FINDING THE INDEX OF REFRACTION - WebAssign

FINDING THE INDEX OF REFRACTION - WebAssign Name: Book: Period: Due Date: Lab Partners: FINDING THE INDEX OF REFRACTION - WebAssign Purpose: The theme in this lab is the interaction between light and matter. Matter and light seem very different

More information

Geometric theory of inversion and seismic imaging II: INVERSION + DATUMING + STATIC + ENHANCEMENT. August Lau and Chuan Yin.

Geometric theory of inversion and seismic imaging II: INVERSION + DATUMING + STATIC + ENHANCEMENT. August Lau and Chuan Yin. Geometric theory of inversion and seismic imaging II: INVERSION + DATUMING + STATIC + ENHANCEMENT August Lau and Chuan Yin January 6, 2017 Abstract The goal of seismic processing is to convert input data

More information

e. \ 2-99 Calculation ofray Amplitudes: We seek a computationally tractable method to estimate ray

e. \ 2-99 Calculation ofray Amplitudes: We seek a computationally tractable method to estimate ray 2-99 e. \ Michael C. Fehler, Steven T. Hildebrand, Lianjie Huang, and Douglas Alde MS 044; Los Alamos National Laboratiory; Los Alamos, NM 87545 USA Summary: Kirchhoff migration using ray tracing travel

More information

Velocity Model Sampling and Interpolation: The Gradient Problem

Velocity Model Sampling and Interpolation: The Gradient Problem January 11, 2013 1 2 3 4 5 Kirchhoff migration uses raytracing equations for prestack depth migration. d x dt = v( x)2 p d p dt = v( x) v( x) These equations estimate the paths of seismic energy through

More information

PS User Guide Series Dispersion Image Generation

PS User Guide Series Dispersion Image Generation PS User Guide Series 2015 Dispersion Image Generation Prepared By Choon B. Park, Ph.D. January 2015 Table of Contents Page 1. Overview 2 2. Importing Input Data 3 3. Main Dialog 4 4. Frequency/Velocity

More information

Summary. Introduction

Summary. Introduction Yaofeng He and Ru-Shan Wu Institute of Geophysics and Planetary Physics, University of California, Santa Cruz, C 9564 Summary We propose an approach to perform migration and imaging using secondary scattered

More information

Wave-equation MVA applied to 4-D seismic monitoring

Wave-equation MVA applied to 4-D seismic monitoring Stanford Exploration Project, Report 112, November 11, 2002, pages 15 21 Short Note Wave-equation MVA applied to 4-D seismic monitoring Paul Sava, John Etgen, and Leon Thomsen 1 INTRODUCTION 4-D seismic

More information

The geometry of reflection and refraction Wave conversion and reflection coefficient

The geometry of reflection and refraction Wave conversion and reflection coefficient 7.2.2 Reflection and Refraction The geometry of reflection and refraction Wave conversion and reflection coefficient The geometry of reflection and refraction A wave incident on a boundary separating two

More information

Progress Report on: Interferometric Interpolation of 3D SSP Data

Progress Report on: Interferometric Interpolation of 3D SSP Data Progress Report on: Interferometric Interpolation of 3D SSP Data Sherif M. Hanafy ABSTRACT We present the theory and numerical results for interferometrically interpolating and extrapolating 3D marine

More information

Effects of multi-scale velocity heterogeneities on wave-equation migration Yong Ma and Paul Sava, Center for Wave Phenomena, Colorado School of Mines

Effects of multi-scale velocity heterogeneities on wave-equation migration Yong Ma and Paul Sava, Center for Wave Phenomena, Colorado School of Mines Effects of multi-scale velocity heterogeneities on wave-equation migration Yong Ma and Paul Sava, Center for Wave Phenomena, Colorado School of Mines SUMMARY Velocity models used for wavefield-based seismic

More information

Motion Planning. Howie CHoset

Motion Planning. Howie CHoset Motion Planning Howie CHoset What is Motion Planning? What is Motion Planning? Determining where to go Overview The Basics Motion Planning Statement The World and Robot Configuration Space Metrics Algorithms

More information

An imaging technique for subsurface faults using Teleseismic-Wave Records II Improvement in the detectability of subsurface faults

An imaging technique for subsurface faults using Teleseismic-Wave Records II Improvement in the detectability of subsurface faults Earth Planets Space, 52, 3 11, 2000 An imaging technique for subsurface faults using Teleseismic-Wave Records II Improvement in the detectability of subsurface faults Takumi Murakoshi 1, Hiroshi Takenaka

More information

Plane Wave Imaging Using Phased Array Arno Volker 1

Plane Wave Imaging Using Phased Array Arno Volker 1 11th European Conference on Non-Destructive Testing (ECNDT 2014), October 6-10, 2014, Prague, Czech Republic More Info at Open Access Database www.ndt.net/?id=16409 Plane Wave Imaging Using Phased Array

More information

Stanford Exploration Project, Report 111, June 9, 2002, pages INTRODUCTION THEORY

Stanford Exploration Project, Report 111, June 9, 2002, pages INTRODUCTION THEORY Stanford Exploration Project, Report 111, June 9, 2002, pages 231 238 Short Note Speeding up wave equation migration Robert G. Clapp 1 INTRODUCTION Wave equation migration is gaining prominence over Kirchhoff

More information

Bellman s Escape Problem for Convex Polygons

Bellman s Escape Problem for Convex Polygons Bellman s Escape Problem for Convex Polygons Philip Gibbs philegibbs@gmail.com Abstract: Bellman s challenge to find the shortest path to escape from a forest of known shape is notoriously difficult. Apart

More information

Tutorial on Multichannel Analysis of Surface Wave (MASW) in the RadExPro Software

Tutorial on Multichannel Analysis of Surface Wave (MASW) in the RadExPro Software Tutorial on Multichannel Analysis of Surface Wave (MASW) in the RadExPro Software (Revision of 23.11.2016) DECO Geophysical SC Moscow State University Science Park 1-77 Leninskie Gory Moscow 119992, Russia

More information

1D internal multiple prediction in a multidimensional world: errors and recommendations

1D internal multiple prediction in a multidimensional world: errors and recommendations 1D internal multiple prediction 1D internal multiple prediction in a multidimensional world: errors and recommendations Pan Pan and Kris Innanen ABSTRACT Internal multiples are more difficult to estimate

More information

P052 3D Modeling of Acoustic Green's Function in Layered Media with Diffracting Edges

P052 3D Modeling of Acoustic Green's Function in Layered Media with Diffracting Edges P052 3D Modeling of Acoustic Green's Function in Layered Media with Diffracting Edges M. Ayzenberg* (Norwegian University of Science and Technology, A. Aizenberg (Institute of Petroleum Geology and Geophysics,

More information

On the Use of Fast Marching Algorithms for Shortest Path Distance Calculation

On the Use of Fast Marching Algorithms for Shortest Path Distance Calculation On the Use of Fast Marching Algorithms for Shortest Path Distance Calculation J.B. Boisvert There are many uses to shortest path algorithms; in past CCG papers the shortest path algorithm has been used

More information

Notes on Minimum Spanning Trees. Red Rule: Given a cycle containing no red edges, select a maximum uncolored edge on the cycle, and color it red.

Notes on Minimum Spanning Trees. Red Rule: Given a cycle containing no red edges, select a maximum uncolored edge on the cycle, and color it red. COS 521 Fall 2009 Notes on Minimum Spanning Trees 1. The Generic Greedy Algorithm The generic greedy algorithm finds a minimum spanning tree (MST) by an edge-coloring process. Initially all edges are uncolored.

More information

A Singular Example for the Averaged Mean Curvature Flow

A Singular Example for the Averaged Mean Curvature Flow To appear in Experimental Mathematics Preprint Vol. No. () pp. 3 7 February 9, A Singular Example for the Averaged Mean Curvature Flow Uwe F. Mayer Abstract An embedded curve is presented which under numerical

More information

Considerations in 3D depth-specific P-S survey design

Considerations in 3D depth-specific P-S survey design Considerations in 3D depth-specific P-S survey design Don C. Lawton and Peter W. Cary 3D P-S survey design ABSTRACT A new sparse-shot design for 3D P-S surveys is introduced. In the sparse shot design

More information

Physics 202, Lecture 23

Physics 202, Lecture 23 Physics 202, Lecture 23 Today s Topics Lights and Laws of Geometric Optics Nature of Light Reflection and Refraction Law of Reflection Law of Refraction Index of Reflection, Snell s Law Total Internal

More information

Sound-speed tomography using first-arrival transmission ultrasound for a ring array

Sound-speed tomography using first-arrival transmission ultrasound for a ring array Sound-speed tomography using first-arrival transmission ultrasound for a ring array Youli Quan* a,b and Lianjie Huang b a Department of Geophysics, Stanford University, Stanford, CA 9435-2215 b Mail Stop

More information

GEOPHYSICS. Time to depth conversion and seismic velocity estimation using time migration velocities

GEOPHYSICS. Time to depth conversion and seismic velocity estimation using time migration velocities Time to depth conversion and seismic velocity estimation using time migration velocities Journal: Manuscript ID: Manuscript Type: Date Submitted by the Author: Geophysics draft Letters n/a Complete List

More information

Downward propagation: using true near-surface velocity model, one-way wave- equation extrapolator:

Downward propagation: using true near-surface velocity model, one-way wave- equation extrapolator: Class 9: Beyond Seismic Statics Corrections Wed, Oct 7, 2009 Dynamic corrections: wave-equation and wavefield datuming methods Skipping the near-surface by interferometry (no model needed) Common-Focus

More information

Kinematics of Machines Prof. A. K. Mallik Department of Mechanical Engineering Indian Institute of Technology, Kanpur. Module - 3 Lecture - 1

Kinematics of Machines Prof. A. K. Mallik Department of Mechanical Engineering Indian Institute of Technology, Kanpur. Module - 3 Lecture - 1 Kinematics of Machines Prof. A. K. Mallik Department of Mechanical Engineering Indian Institute of Technology, Kanpur Module - 3 Lecture - 1 In an earlier lecture, we have already mentioned that there

More information

0580 MATHEMATICS. 0580/31 Paper 3 (Core), maximum raw mark 104

0580 MATHEMATICS. 0580/31 Paper 3 (Core), maximum raw mark 104 CAMBRIDGE INTERNATIONAL EXAMINATIONS International General Certificate of Secondary Education www.xtremepapers.com MARK SCHEME for the May/June 2014 series 0580 MATHEMATICS 0580/31 Paper 3 (Core), maximum

More information

Summary. Introduction

Summary. Introduction Chris Davison*, Andrew Ratcliffe, Sergio Grion (CGGeritas), Rodney Johnston, Carlos Duque, Jeremy Neep, Musa Maharramov (BP). Summary Azimuthal velocity models for HTI (Horizontal Transverse Isotropy)

More information

Data Structure. IBPS SO (IT- Officer) Exam 2017

Data Structure. IBPS SO (IT- Officer) Exam 2017 Data Structure IBPS SO (IT- Officer) Exam 2017 Data Structure: In computer science, a data structure is a way of storing and organizing data in a computer s memory so that it can be used efficiently. Data

More information