Difference between revisions of "La Habra Simulations on Titan"

From SCECpedia
Jump to navigationJump to search
Line 131: Line 131:
 
*STRIKE = 239
 
*STRIKE = 239
 
*RAKE = 38
 
*RAKE = 38
*DIP = 79 (typo? should be 70?)
+
*DIP = 70
 
*LAT_TOP_CENTER = 33.9119
 
*LAT_TOP_CENTER = 33.9119
 
*LON_TOP_CENTER = -117.940
 
*LON_TOP_CENTER = -117.940

Revision as of 18:40, 9 July 2014

Overview

A series of simulations modelling the La Habra 5.1 event will be performed on ORNL Titan with in order to address these scientific and computational goals:

  • Demonstrate that Hercules-GPU software is in production status and evaluate its performance versus the CPU version
  • Evaluate goodness of fit for synthetic waveforms versus observed at 1.0 Hz, using Po Chen's updated CVM-S4 velocity model as the material properties source for the simulation.
  • [Optionally] Validate the simulation results generated by Hercules-GPU and AWP-GPU


Solver Parameters

Parameter Value Notes
Frequency 1.0 Hz
Simulation Length 100 s
Delta T 0.005
Plane Output Resolution 250m
I/O Print Rate every 10 steps
Station List LH_stations.txt Merged from En-Jui's 0.2Hz station list and Ricardo's CH master list
Software Version Hercules-GPU commit c95f1ab017e5f93613a0e3e2c6b5a849eb4117ad GPU codes, no frequency dependent Q


Simulation Box / Velocity Model

Parameter Value Notes
Dimensions (km) 180 x 135 x 61.875
Bounding Box (LL) -119.288842 34.120549, -118.354016 35.061096, -116.846030 34.025873, -117.780976 33.096503
UCVM Version 13.9.0 No heterogeneities
Velocity Model Versions CVM-S4, CVM-S4 v26 CVM-S5?
Miniumum Vs 200 m/s
Samples per wavelength 8
Hercules Etree ch-cvms400-100-4hz-200ms.e, ch-cvms426-223-4hz-200ms.e CVM-S5 etree?


Sources

The simulation will be repeated for two point sources and one extended source.

Point Source Parameters

Parameter Value Notes
Event Name La Habra
Origin Time 2014/03/29 04:09:42.97 Source: En-Jui
Origin Location -117.930; 33.922; 5.0km Source: En-Jui
Strike/Dip/Rake 1 134/55/155 Source: En-Jui, initial
Strike/Dip/Rake 2 239/70/38 Source: En-Jui, based on review of aftershock sequence
Rise Time 0.75 s


SRC Format La Habra Source

  • MAGNITUDE = 5.12
  • FAULT_LENGTH = 5.0
  • FAULT_WIDTH = 2.25
  • DEPTH_TO_TOP = 4.7125
  • STRIKE = 239
  • RAKE = 38
  • DIP = 70
  • LAT_TOP_CENTER = 33.9119
  • LON_TOP_CENTER = -117.940
  • HYPO_ALONG_STK = 0.0
  • HYPO_DOWN_DIP = 1.125
  • DLEN = 0.1
  • DWID = 0.1

Volume

The four corners are:

  • -119.288842 34.120549
  • -118.354016 35.061096
  • -116.846030 34.025873
  • -117.780976 33.096503

And the box has size

  • 180,000 m x 135,000 m x 61,875 m

Runs

Hercules Runs

Run Number Etree Source Notes
1 ch-cvms400-100-4hz-200ms.e point-ricardo-134-55-155
2 ch-cvms426-223-4hz-200ms.e point-ricardo-134-55-155
3 ch-cvms400-100-4hz-200ms.e point-ricardo-239-70-38
4 ch-cvms426-223-4hz-200ms.e point-ricardo-239-70-38
5 ch-cvms400-100-4hz-200ms.e extended-ricardo
6 ch-cvms426-223-4hz-200ms.e extended-ricardo


BBP Run


La Habra Earthquake Observations

La Habra References


References

  1. Bielak, J., H. Karaoglu, and R. Taborda, 2011. Memory-efficient displacement-based internal friction for wave propagation simulation, Geophysics, 76(6):T131-T145.
  2. Taborda, R., Lopez, J., Karaoglu, H., Urbanic, J., and Bielak, J. (2010). Speeding up finite element wave propagation for large-scale earthquake simulations. Technical Report CMU-PDL-10-109, Carnegie Mellon University, Parallel Data Lab.
  3. Tu, T., Yu, H., Ramírez-Guzmán, L., Bielak, J., Ghattas, O., Ma, K.-L., & O’Hallaron, D.R., 2006. From mesh generation to scientific visualization: an end-to-end approach to parallel supercomputing, in Proceedings of the 2006 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis, p. 15, IEEE Computer Society, Tampa, Florida.