PhD

Objective

Develop, apply and test machine learning (ML) techniques for training and labeling the increasing number of seismological datasets in a complete earthquake detection workflow, aiming to improve real-time earthquake monitoring capabilities, including the recognition of micro-seismic patterns leading to main shocks.

The PhD Workspace has the following file organization structure.

├── data
│   ├── processed
│   │   ├── 20230601
│   │   │   ├── 4P
│   │   │   │   └── IT09A
│   │   │   ├── CH
│   │   │   │   ├── BERNI
│   │   │   │   ├── ...
│   │   │   │   └── SZER
│   │   │   ├── ...
│   │   │   └── OX
│   │   │       └── MLN
│   │   ├── ...
│   │   └── 20230603
│   │       └── IV
│   │           └── FVI
│   ├── test
│   │   ├── annotated
│   │   ├── classified
│   │   └── waveforms
│   └── waveforms
├── doc
│   ├── Doc
│   └── References
├── img
├── src
│   ├── __init__.py
│   └── AdriaArray.py
└── test
    ├── __init__.py
    └── test_AdriaArray.py

In the folder src you will find the file AdriaArray.py. This file is able to be excuted pointing to any required directions. The program AdriaArray.py is an invasive program, in the sense that creates files (by default) at the same level directory as the folder containing the raw waveforms. Therefore, please consider available memory space, especially when working with limited external devices.

An axample of this behaviour can be seen in the tree above. In the folder data, there exists the folder waveforms which contains the files to be analyzed by AdriaArray pipeline. The program will create the folders 'processed', 'classified' and 'annotated' at the same level as waveforms.

The project contains several manually implemented tests to provide a feedback of the reliability of the program, as well a framework to understand where a potential bug could be located during user execution of the program. In order to execute these tests, is as simple as typing the following command in the terminal:

% make testing

which will launch all the tests to establish the reliability known up to date. The data saved in the test are either randomly sampled data or they provide specific data to which we can cuantitatively measure the results obtained by the program.

AdriaArray

AdriaArray allows the user to either customize by predefining the default behaviour or specify all the variables on demand.

TODO: Consider file parsing

AdriaArray allows the following commands on demand:

• -h or --help: show help message and exit
• -C or --channel: The user requests a sequence of channels to be analyzed. If file is not available, then a key must be provided in order to download the data.
• -D or --dates: The user must define the initial date and the last date (inclusive) (i.e. range of dates) to be analyzed.
• -G or --groups: Analyze the data based on groups which are categorical list of files which meet the criteria.
• -K or --key: Key to have access to download the data from server.
• -M or --models: Select a specific Machine Learning based model.
• -N or --network: The user requests a sequence of networks to be analyzed. If file is not available, then a key must be provided in order to download the data.
• -S or --station: The user requests a sequence of stations to be analyzed. If file is not available, then a key must be provided in order to download the data.
• -T or --train: The user establishes the code to execute in TRAINING mode.
• -W or --weights: Select a specific pretrained weights for the selected Machine Learning based model. WARNING: Weights which are not available for the selected models will not be considered.
• -d or --directory: The user defines the directory in which the raw data are stored.
• -v or --verbose: The user establishes the maximum level of communication of the program towards the user. It is useful if the user wants to interact with the output of the program.

How to get started?

An easy way to get started is by executing the following command:

% python src/AdriaArray.py -v --directory data/test/waveforms

This will run the few test examples of data we have considered worthwhile saving for testing purpouses. It will print all of the messages possible of the execution of the program and the user will be able to pause and interact indefenetely to analyze the output of the program. In order to continue to the next results the user must close the graph plot such that the program may continue executing.

Next steps

The following command:

% python src/AdriaArray.py -v -D 19980101 19980110 --directory path/to/waveforms -M PhaseNet EQTransformer -W instance stead original

will try to search the existance of path/to/waveforms during the dates 1998/01/01 and 1998/01/10 and apply all the possible combinations between the models Phasenet, EQTransformer and the pretrained weights instance, stead, original

Good luck!