This repository provides scripts to calibrate, apply and analyze correction functions for magnitude estimation. The correction functions consist of three parts:
- station bias
- 2D distance and depth correction
- 3D source correction (station specific)
It further provides functions to reduce uncertainty through multi-feature estimation using boosting tree regression.
The scripts use Python>=3.5. Please install python requirements with pip install -r requirements.txt.
Additionally the script relies on the Gurobi Optimizer for quadratic optimization. Gurobi provides free Academic Licenses (http://www.gurobi.com/academia/for-universities). Please also install the additional Gurobi python library.
For details on usage please consult the Jupyter notebook and refer to the documentation in the code.
@article{munchmeyer2019,
author = {Münchmeyer, Jannes and Bindi, Dino and Sippl, Christian and Leser, Ulf and Tilmann, Frederik},
title = "{Low uncertainty multi-feature magnitude estimation with 3D corrections and boosting tree regression: application to North Chile}",
journal = {Geophysical Journal International},
year = {2019},
doi = {10.1093/gji/ggz416}
}
This software is licensed under GPL v. 3.0. For the calibration of magnitude scales, this software interacts with the Gurobi Solver Software via the gurobipy API; Gurobi Solver Software and the gurobipy API are only available under the End-User License Agreement as provided by Gurobi Optimization, LLC, ("Gurobi EULA"). You need your own license under the Gurobi EULA to use this functionality.