📄 readme
字号:
This is a README to the FATOMO code (Ver. Oct., 2000) to give a brief overview how to use fatomo describing the different input & output files. Dr. St. Husen Inst. of Geophysics ETH-Hoenggerberg 8057 Zurich Switzerland PHONE: ++41-1-6332622 EMAIL: stephan@tomo.ig.erdw.ethz.ch Jan., 2001 GENERAL REMARKS FATOMO inverts for 3D Vp-velocity structure using the fat ray concept (Husen & Kissling, Phys. Earth Planet. Int., 123, 127-147, 2001. Travel times are calculated using a FD-scheme proposed by Podvin & Lecomte (Geophys. J. Int., 105, 271-284, 1991). Fat rays resembling Fresnel volumes are used to compute model partial derivatives. For more information on the fat ray concept please visit the webpage: www.sg.geophys.ethz.ch/aes/husen FATOMO solves the hypocentral part of the coupled hypocenter-velocity problem using a grid search algorithm. The grid search algorithm implemented in FATOMO requires well-locatable events and hypocenter locations close to the true location, f.e. obtained by relocating the events with a minimum 1D model with including station corrections. The grid search is performed over two grids: First on the coarse, seismic grid and second on the fine numerical grid. Search radius for the search on the seismic grid is given in the CNTL file in x,y,z direction and in km. The initial search radius show include several seismic grid nodes in each direction to ensure that the global minimum is included. For the second search on the numerical grid, a box is set up with its center at the location of the lowest RMS value of the initial search with sided defined by the surrounding seismic grid nodes. For each hypocenter location the 2 sigma interval (corresponding to the 95% confidence interval) is given in x,y,z direction. A few words to memory handling of FATOMO: Computation of the fat rays requires both traveltime fields of the source and the receiver to be computed. Traveltime fields of the receivers are calculated ahead of each inversion and stored on disk (files: recXXX.tmp) to reduce RAM allocation of FATOMO. To reduce the burden of loading traveltime fields from disk, traveltime fields for all stations observing the event are read from disk prior to relocation resp. fat ray computation. These traveltime fields are stored in the 2-D array "tt_obs", which is of the dimension maxobs x maxblocks (maxobs: max no. obs per event; maxblocks: max no. cells of the FD grid). Therefore memory allocated by FATOMO depends on the size of the velocity model (max. dimension in x,y,z km) and on the maximum numbers of observations per event. The included testcase allocates 260 MBytes of RAM (maxobs=25, maxblocks=200x200x100). If your data set is larger, you must modify the corresponding variables in the include file "fatomo_common.inc". Parts of the code, i.e. input routines and inversion routines, were taken from the SIMULPS13Q code. MODEL SET UP Fatomo uses three different grids in the inversion. The SEISMIC grid consists of grid nodes at intersecting lines with linear interpolation in between as in SIMULPS13Q. Node spacing can be uneven and node positions will be given relative to the short distance conversion origin, where east and south are NEGATIVE. The INVERSION grid consist of blocks, which are defined by their edges in x,y,z. Thereby,two restrictions have to be met: A possible block edge must be located at the midpoint between two seismic grid points and spacing must be a MULTIPLE of the numerical gridspacing. By omitting block edges it is possible to combine several seismic grid nodes to one inversion block. It is possible to vary the number of inversion block edges, i.e. the number of inversion blocks, for each layer. This allows, for example, larger inversion blocks with increasing depth to account for decreasing ray coverage. Inversion block edges must be given for each layer in the MOD file (see below). Similar to SIMULPS13Q inversion blocks can be individually a priori fixed during the inversion. Inversion blocks to be fixed are specified by their i,j,k-index (i.e. number in x,y,z direction). First block which can be fixed is 1,1,1 located in the lower left corner of the first inversion layer. The NUMERICAL grid is used by the FD modeling to compute synthetic travel times. Since the FD-scheme uses a left-handed coordinate system, origin of the numerical grid is at the lower left corner (relative to the xy-plane) of the seismic grid. Internally, all station & event coordinates are transformed relative to this origin. Grid spacing of the numerical grid is given in the control file and should be as fine as possible. Velocities are linearly extrapolated from the coarser seismic grid. NOTE: Due to the definition within the Podvin and Lecomte FD-code velocities in those cells with the highest number, i.e. located at right and upper boundary of the model, are set to infinity. Therefore design the seismic grid at least one grid spacing larger than required by the station and event distribution. Seismic grid nodes and inversion block boundaries are given in the model inputfile 'MOD' (see below). Its structure is equivalent to that of SIMULSPS13Q with one EXCEPTION: the position of grid nodes resp. block edges in x-direction must be arranged in DECREASING order, because the origin of the numerical grid is at the lower left. In y- and z-direction positions are arranged in INCREASING order. Counting of seismic grid nodes resp. inversion blocks start at the origin of the numerical grid, i.e. LOWER LEFT corner of the first layer. This is contrary to SIMULPS13Q where the first grid node is located at the lower right corner. The MOD-file starts with the number of block edges (inv_nz) of the inversion grid in z-direction (free format) followed by their position. For each layer of the inversion grid (inv_nz-1) the layer number and the number of inversion block edges in x and y-direction are given next followed by their position (one line for each direction). First and last inversion block edges define the model dimension in x,y,z. The seimic grid is given by the number of seismic grid nodes (including seismic grid nodes at the model boundaries) in x,y,z direction followed by their position (one line for each direction). Following the seismic grid is a list of inversion blocks which are a priori fixed during the inversion. The list terminates with ' 0 0 0'. Finally, the initial velocities are given for each grid node starting with the UPPER LEFT corner of the top layer. X-direction is the fastest index. If the model consists of more than 20 grid nodes in x-direction, a new line must be open for the next 20 grid nodes. Format of the MOD-file is a free-format (*). Position of inversion block edges and seismic grid nodes are given relative to origin of short distance origin (E and S are NEGATIVE). Example MOD file (see Testcase): 12 # inversion grid -4.0 1.0 3.0 4.5 7.5 12.5 17.5 22.5 27.5 32.5 37.5 40.0 1 15 15 80.0 72.5 62.5 57.5 52.5 47.5 42.5 37.5 32.5 27.5 22.5 17.5 12.5 5.0 0.0 0.0 5.0 12.5 17.5 22.5 27.5 32.5 37.5 42.5 47.5 52.5 57.5 62.5 72.5 80.0 2 15 15 80.0 72.5 62.5 57.5 52.5 47.5 42.5 37.5 32.5 27.5 22.5 17.5 12.5 5.0 0.0 0.0 5.0 12.5 17.5 22.5 27.5 32.5 37.5 42.5 47.5 52.5 57.5 62.5 72.5 80.0 ..... 4 8 8 80.0 62.5 52.5 42.5 32.5 22.5 12.5 0.0 0.0 12.5 22.5 32.5 42.5 52.5 62.5 80.0 ..... 10 15 15 80.0 72.5 62.5 57.5 52.5 47.5 42.5 37.5 32.5 27.5 22.5 17.5 12.5 5.0 0.0 0.0 5.0 12.5 17.5 22.5 27.5 32.5 37.5 42.5 47.5 52.5 57.5 62.5 72.5 80.0 14 14 11 # seismic grid 80.0 65.0 60.0 55.0 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 0.0 0.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 55.0 60.0 65.0 80.0 -4.0 0.0 2.0 4.0 5.0 10.0 15.0 20.0 25.0 30.0 40.0 1 1 1 # inv. blocks a priori fixed 1 1 4 1 1 5 1 1 6 0 0 0 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 ...... CONTROL FILE (file CNTL) Control parameters are given via the file CNTL. Comments start with
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -