Skip to content

simpeg/aurora

Repository files navigation

AURORA


Aurora is an open-source package that robustly estimates single station and remote reference electromagnetic transfer functions (TFs) from magnetotelluric (MT) time series. Aurora is part of an open-source processing workflow that leverages the self-describing data container MTH5, which in turn leverages the general mt-metadata framework to manage metadata. These pre-existing packages simplify the processing by providing managed data structures, transfer functions to be generated with only a few lines of code. The processing depends on two inputs -- a table defining the data to use for TF estimation, and a JSON file specifying the processing parameters, both of which are generated automatically, and can be modified if desired. Output TFs are returned as mt-metadata objects, and can be exported to a variety of common formats for plotting, modeling and inversion.

Key Features

  • Tabular data indexing and management (Pandas dataframes),
  • Dictionary-like processing parameters configuration
  • Programmatic or manual editing of inputs
  • Largely automated workflow

Documentation for the Aurora project can be found at http://simpeg.xyz/aurora/

Installation

Suggest using PyPi as the default repository to install from

pip install aurora

Can use Conda but that is not updated as often

conda -c conda-forge install aurora

General Work Flow

  1. Convert raw time series data to MTH5 format, see MTH5 Documentation and Examples.
  2. Understand the time series data and which runs to process for local station RunSummary.
  3. Choose remote reference station KernelDataset.
  4. Create a recipe for how the data will be processed Config.
  5. Estimate transfer function process_mth5 and out put as a mt_metadata.transfer_function.core.TF object which can output [ EMTFXML | EDI | ZMM | ZSS | ZRR ] files.