Deformation Analysis and Modeling for the 2011 Tohoku-Oki Earthquake
On March 11th, 2011, a giant megathrust happened off the Pacific coast of
Tohoku, Japan. This earthquake was the most powerful earthquake ever recorded to
have hit Japan and causes extensive co- and post-seismic deformation, which
is recorded by more than one thousands GPS stations in and around Japan.
This software package is designed for the analyze of GPS time series that are related with
this earthquake and do the linear and non-linear modeling to infer earthquake
parameters and the rheological parameters of the Earth. Specifically, this
package contains the following parts:
1) GPS time series management (sqlite3 database);
2) GPS time series modeling - Linear modeling of preseismic GPS time series;
3) GPS time series modeling - Nonlinear inversion of coseismic slip and afterslip (solve for more than 10, 000 parameters);
4) Results analysis and report.
All libraries developed for the project can be found under lib/ directory.
All the unit tests of these libraries are under test/.
Other directories are for specific experiments or tasks.
All libraries developed for this project can be found under lib/ directory.
All the linear and nonlinear analysis are implemented in this library.
-
lib/viscojapan/tsana/ Linear analysis R wrapper for linear regression analysis.
-
lib/viscojapan/ Nonlinear analysis – The core is the method of convex optimization provided by an open source Python package called cvxopt. There are at least three layers are built upon this computational core to handle different situations of my projects.
Multiprocessing with Python
I designed and implemented a library called Dynamic POOL (dpool),
where the number of running processes is dynamically managed so that
when the system load from other users is heavy, the number of
processes is reduced, and vice versus.
A Python wrapper for Latex to combine subplots.
A wrapper for GMT for geographic plotting.
- Python3 : Python is the main language used to write this package.
- R : R is integrated with Python to do linear regression analysis of GPS time series.
- C/C++ : Some parts of Python are written with C/C++ to enhance performance.
- numpy : for scientific computation
- scipy : for scientific computation
- matplotlib : for 2D plotting
- basemap : for 2D geographic plotting
- rpy2 : R python wrapper
- h5py : For results storage and efficient large data accessing
- cvxopt : Python library for convex optimization
- pyproj : Performs cartographic transformations and geodetic computations
- simplekml : generate kml (or kmz) for displaying data on google map/earth
For any questions, please contact the author:
Yang Zhang (zy31415@gmail.com)
University of Nevada, Reno