This branch is the code base for running a dash app for visualising and interpreting airborne electromagnetic data. A feature of this app is that it is able to create visualisations of a bulk conductivity model ensemble generated using a stochastic AEM inversion code. The plots that are generated with this app are interactive and can be used for interpreting stratigraphic features with predictive uncertainty.
This package has many dependencies and we use anaconda to manage them. The following conda commands create an environment that should run the app no problem.
conda create -n AEM_interp python=3.7 conda activate AEM_interp
conda install -c conda-forge numpy pandas netCDF4 shapely geopandas matplotlib scipy dash conda install -c conda-forge xarray dask bottleneck conda install -c scikit-learn conda install -c plotly pip install affine pip install pyyaml
pip install -e
The first step in running the dash app is to prepare the data and point the app to it. We assume a user has already run AEM inversions using both a deterministic and/ or stochastic code. We use contractor ga_aem (https://github.com/GeoscienceAustralia/ga-aem) for running these inversions on the National Computational Infrastructure (NCI).
NetCDF is a convenient format for storing and handling large point, grid or line data. To convert our AEM line data to netcdf we use the geophys utils package developed at GA: https://github.com/GeoscienceAustralia/geophys_utils/tree/master/geophys_utils . This package itself has a host of dependencies and we use a seperate conda environment. Scripts for converting the EM data and deterministic inversion can be found in the "utils/conversion" directory.
The probabilistic inversions are handles differently. AEM soundings are inverted on a point by point basis, with each inversion creating a single netcdf file. The "utils/conversion/concatenate_netcdf_file.py" script demonstrates how to compile the netcdf files into a single file containing just the important information. We rely very heavily on consistent field names for this app so we highly recommend a user familiarising themselves with the "utils/conversion/netcdf_settings.yaml" file.
Once all the data are prepared, they should be copied into a single directory. Use the "workflow/interpretation_config.yaml" file to set the path to this directory. To avoid expensive gridding of AEM line data onto sections/ grids, we preprocess the data and save the 'rasterised' sections as xarray files. The gridding parameters and path are set using the interpretation_config.yanl file. More information on how this works can be found in the wiki.
Once the data is prepared and the settings are satisfactory, run the app (app.py) using thw python from the conda environment. If gridding sections, the app may take a while to run. Once the preprocessing is done, the terminal should return a message similar to the following:
Dash is running on http://127.0.0.1:8050/
- Serving Flask app "app" (lazy loading)
- Environment: production WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
- Debug mode: off
- Running on http://127.0.0.1:8050/ (Press CTRL+C to quit)
If all has worked, the app can be opened by pasting the url into your browser (we have tested on firefox and chrome). Happy interpreting :)