CFS_Calculation

This Repo contains package for CFS calculation and ROC analysis as published in https://doi.org/10.1029/2020JB019553

Is Coulomb stress the best choice for aftershock forecasting?

This is an archive of the NRC project I6 part 1

Author: Shubham Sharma
Supervisors: Sebastian hainzl, Gert Zöller and Matthias Holschneider

Archive structure

graph TD;
A{"NRC-I6-1_CFS <br/> (Parent directory)"} --> B("data <br/> (Input and output data)");
A --> C("codes <br/> (shell, python, fortran)");
A --> D("docs <br/> (literature & manuscripts)");
A --> E("example <br/> (example data to test codes)");

Loading

Directory structure

NRC-I6-1_CFS
|   README.md
|
|___example
|   |___figs
|   |   |___stress
|   |   |___ROC
|   |
|   |___workingData
|   |   catalog.dat
|   |   events.txt
|   |
|   |   |___PSCMP-INPUTFILES
|   |   |   |___okada
|   |   |   |___GF
|   |   |
|   |   |___PSGRN-INPUTFILES
|   |   |___psgrn+pscmp_out
|   |   |   |___pscmp_out
|   |   |   |   |___okada
|   |   |   |   |___GF
|   |   |   |
|   |   |   |___psgrn_out
|   |   |   |
|   |   |   |___tmp_out
|   |   |___srcmod
|   |   |___interData
|   |   |   |___ASFiles
|   |   |   |___polys
|   |   |
|   |   |___ROCData
|   |   |   |___ROCout
|   |   |   |   |___MAS0
|   |   |   |   |___MAS
|   |   |   |   |___OOP
|   |   |   |   |___VM
|   |   |   |   |___MS
|   |   |   |   |___VMS
|   |   |
|   |   |___stressData
|   |   |   |___MAS0
|   |   |   |___MAS
|   |   |   |___OOP
|   |   |   |___VM
|   |   |   |___MS
|   |   |   |___VMS
|   |
|___codes
|   |___packages
|   |   |___PSGRN+PSCMP
|   |   |___CRUST_2.0
|   |
|   |___1_preProcessing
|   |___2_genStressTensors
|   |___3_ROCAnalysis
|   |___4_plotting
|
|___docs
|   Sharma_IS-CFS-best-choice_2020_JGR-Solid-Earth.pdf

Directory contents

NRC-I6-1_CFS: Parent directory

• rawData: Directory (can be created in parent directory) to store ISC Catalog in pickle format, which can be downloaded from ISC website.

• example: an example folder where all the codes can be tested
  |---workingData: contains pre-processed data and results

• codes: contains all codes related to this work
  |---packages: external packages used
  |---1_preProcessing: all pre-processing code for CFS calculation using PSGRN+PSCMP
  |---2_genStressTensors a shell script to run PSGRN+PSCMP codes
  |---3_ROCAnalysis ROC analysis scripts to calculate ROC values
  |---4_plotting Plotting scripts to plot stress maps and ROC curves

• docs: Publications and related material

Working steps

NOTE:

Please create the directories mentioned in directory tree [example] before proceeding for calculations.

1. Pre-processing

   • scripts used:

     • fsp2PSGRN.py [Prepare SRCMOD files to be used as input in PSGRN]
     • fsp2PSCMP.py [Prepare SRCMOD files (okada or GF) to be used as input in PSCMP]

   • inputs:

     • catalog.dat and SRCMOD .fsp files

   • Working steps:

     • goto directory example and run

   python ../codes/1_preProcessing/fsp2PSGRN.py
   python ../codes/1_preProcessing/fsp2PSCMP.py
   

   • output:
     • output would be saved in workingData/PSGRN-INPUTFILES and workingData/PSCMP-INPUTFILES/[okada/GF]

2. Stress tensors and CFS using PSCMP+PSGRN program

   • scripts used:

     • run_PSGRN+PSCMP.sh [Generate stress tensors using Rongjiang's PSGRN+PSCMP codes]

   • inputs:

     • Inputs for running PSGRN are saved in PSGRN-INPUTFILES
     • Inputs for running PSCMP are saved in PSCMP-INPUTFILES/[okada/GF]

   • working steps:

     • goto directory example and run

   ../codes/2_genStressTensors/run_PSGRN+PSCMP.sh
   

   • outputs:
     • outputs are saved in workingData/psgrn+pscmp_out/

3. ROC analysis

   • Standard scripts for calculating ROC curves are given, in order to conduct tests regarding different grid-size, magnitude cut-off and aftershock duration modify parameters in scripts accordingly and produce results.

   • This script also save stress metric values for different stress in workingData/stressOut folder

   • scripts used:

     • calcROC.py [Generate ROC values as well as save stress metric]

   • inputs:

     • stress files saved in pscmp_out
     • srcmod file saved in srcmod
     • event catalog file events.txt
     • aftershock catalog saved as binary isc_rev.pkl

   • working steps:

     • go to directory example and run

   python ../codes/3_ROCAnalysis/calcROC.py
   

   • outputs:
     • ROC values are saved in directory ROCData/ROCout/ for all stress metric
     • stress metric values saved in stressData/

4. Plotting

   • standard scripts for plotting stress maps and ROC curves

   • scripts used:

     • plotStress.py [Plots stress maps for a slip model]
     • plotROC.py [Plots ROC curves for different user inputs]

   • inputs:

     • Stress files saved in workingData/stressData
     • ROC files saved in ROCData/ROCout

   • working steps:

     • go to directory example and run

   python ../codes/4_plotting/plotStress.py
   python ../codes/4_plotting/plotROC.py
   

   • outputs:
     • output figures (stress maps and ROC curves) are saved in figs/[stress/ROC] directory

For any help regarding the programs contact sharma@gfz-potsdam.de