|
| 1 | + |
| 2 | +xnmsyn calculates a normal mode synthetics using normal mode |
| 3 | +summation for given CMTSOLUTION and STATIONS files. |
| 4 | + |
| 5 | +1. INPUT PARAMTERS: |
| 6 | + |
| 7 | +EXAMPLE--- |
| 8 | +/home/lqy/catalogue/data/aniprem808.dk |
| 9 | +/home/lqy/catalogue/data/toroidal.aniprem808.bin |
| 10 | +/home/lqy/catalogue/data/radial.aniprem808.bin |
| 11 | +/home/lqy/catalogue/data/spheroidal.aniprem808.bin |
| 12 | +CMTSOLUTION |
| 13 | +8.0 |
| 14 | +1 3600 |
| 15 | + |
| 16 | +are respectively: model deck file, toroidal catalogue, radial catalogue, |
| 17 | +and spheroidal catalogue, CMTSOLUTION file name, shortest period |
| 18 | +in summation, dt in output time series, and total length of record in |
| 19 | +seconds. |
| 20 | + |
| 21 | +example execution: xnmsyn < in_8 |
| 22 | + |
| 23 | +2. SETUP |
| 24 | + |
| 25 | + Make sure you have STATIONS and SYNT/ in the current directory. |
| 26 | + |
| 27 | +file STATIONS has the format of |
| 28 | + |
| 29 | +2 |
| 30 | +KMI IC 25.1233 102.7400 1975.0 0.0 |
| 31 | +PAS TS 34.1490 -118.1720 0.0 0.0 |
| 32 | + |
| 33 | +first line -- number of stations |
| 34 | +second - last line -- station_name, network_code stalat, stalon, staele, burial |
| 35 | + |
| 36 | + |
| 37 | +CMTSOLUTION files are the Harvard format: |
| 38 | + |
| 39 | +PDE 1994 6 9 0 33 45.40 -13.8200 -67.2500 647.1 6.9 6.8 NORTHERN BOLIVIA |
| 40 | +event name: M060994A |
| 41 | +time shift: 0.0000 |
| 42 | +half duration: 30.0000 |
| 43 | +latitude: -13.8200 |
| 44 | +longitude: -67.2500 |
| 45 | +depth: 647.1000 |
| 46 | +Mrr: -0.759000E+28 |
| 47 | +Mtt: 0.775000E+28 |
| 48 | +Mpp: -0.160000E+27 |
| 49 | +Mrt: -0.250300E+29 |
| 50 | +Mrp: 0.420000E+27 |
| 51 | +Mtp: -0.248000E+28 |
| 52 | + |
| 53 | +3. OUTPUT |
| 54 | + |
| 55 | +OUTPUT will be in SYNT/ with names like: |
| 56 | +PAS.TS.LHE.nmsyn PAS.TS.LHN.nmsyn PAS.TS.LHZ.nmsyn |
| 57 | +and the old version of synthetics with the same name will |
| 58 | +be overwritten, so be sure to rename the old ones if they |
| 59 | +are still useful. |
| 60 | + |
| 61 | +You can type |
| 62 | +xmgr -graph 0 PAS.TS.LHE.nmsyn -graph 1 PAS.TS.LHN.nmsyn -graph 2 PAS.TS.LHZ.nmsyn -rows 3 to see preliminary result. |
| 63 | + |
| 64 | + |
| 65 | +REMARKS: |
| 66 | + |
| 67 | +1. If you want to make synthetics with another mode catalogue, you might |
| 68 | +want to change the parameters in minos.h, including NR and NL. Make sure |
| 69 | +you find out the written format of the catalogue and change correspondingly |
| 70 | +in the nmsyn.f. |
| 71 | + |
| 72 | +2. If you want output to be RTZ component instead of NEZ component, you can |
| 73 | +change parts of write_record.f, or do the rotation with your own program. |
| 74 | + |
| 75 | +3. A minor problem exists for the output synthetics. The synthetics is not |
| 76 | +perfectly zero before P arrivals, and I cannot figure out what causes this. |
| 77 | +You are more than welcomed to find out why. ( Jeroen has a program to remove |
| 78 | +the offset) |
| 79 | + |
| 80 | +4. There are several mode catalogues available for summation: |
| 81 | + |
| 82 | +radial.aniprem808_40s.bin |
| 83 | +toroidal.aniprem808_40s.bin |
| 84 | +spheroidal.aniprem808_40s.bin |
| 85 | +is a 40s mode catalogue for aniprem808 with top 165 layers stored. |
| 86 | + |
| 87 | +radial.aniprem808_8s.bin |
| 88 | +toroidal.aniprem808_8s.bin |
| 89 | +spheroidal.aniprem808_8s.bin |
| 90 | +is a 8s mode catalogue for aniprem808 with top 165 layers stored. |
| 91 | + |
| 92 | +Use the 40s catalogue directly instead of using tmin=40 for the 8s |
| 93 | +catalogue will save you a lot of time. |
| 94 | + |
| 95 | +5. One can compare the output with the output of QmXD in Sun |
| 96 | +workstation, using the following script |
| 97 | + |
| 98 | +#!/usr/bin/tcsh -f |
| 99 | +setenv MYDBS /net/freawaru/export/home2/jtromp/dbs |
| 100 | +/net/freawaru/export/home2/jtromp/bin/QmXD -M \ |
| 101 | +/net/freawaru/export/home1/jtromp/dbs/catalogues/seigsml_prem_an.750.8s \ |
| 102 | +-O 8. -P 6000. -s |
| 103 | + |
| 104 | +The output generally agrees farely well except the problem I described |
| 105 | +earlier. |
| 106 | + |
| 107 | +6. The dt can be chosen as long as 2*dt < Tmin, larger dt will result |
| 108 | +in faster computation. |
| 109 | + |
| 110 | +7. The program was originally written by Jeroen, and I only changed the |
| 111 | +input and output format. |
| 112 | + |
| 113 | + |
| 114 | + |
| 115 | +03/21/02 |
0 commit comments