Atmospheric data Community Toolkit (ACT)

Lightning talk about ACT and how we are trying to bring together members of the atmospheric science community that work with instrument time-series data, do develop an easy to use toolkit for anything from data discovery and reading to corrections and additional processing for retrievals.

Transcript

1. Atmospheric data Community Toolkit
2. Who? What? Why? How? § Adam Theisen (first time at SciPy) § Instrument Operations Manager Atmospheric Radiation Measurement User Facility (ARM) Argonne National Laboratory § Free time == coding! § Work with a lot of data from atmospheric instrumentation; time series x N-dimensional (air temperature, cloud cover, lidars, etc) § Problem – Redundant coding across the community (How many times have grad students had to write code to read data and create visualizations!) – Norm was to share data and not code – “Constants” were not constant across codes § Team from Argonne, University of Oklahoma, and Oak Ridge National Laboratory contributed to developing an open-source python-based community toolkit to: – Reduce duplication of efforts! – Enable easier sharing of code across groups! – Ensure consistency!
3. Atmospheric data Community Toolkit (ACT) § ACT is a python library for working with atmospheric data time series § Structure – xarray Sphinx – Travis CI Coveralls (Code Coverage) – Zenodo (DOI) § Modules – Data Discovery (APIs) I/O – Visualization Quality Control – Corrections Retrievals – General Utilities § Developed to work across programs – ARM NEON – AmeriFlux Array of Things (AOT) – NOAA Argonne Meteorological Tower § Notable additions – Bridge ACT and Python ARM Radar Toolkit (Py-ART) – FFT analysis for radiometer shading problems – Retrieve Infrared temperature from AERI – Correct wind speed and direction for ship motion – Download data from ARM, ASOS, and CropScape using web services Argonne Tower NEON ARM Ameriflux AOT https://github.com/ARM-DOE/ACT 8 contributors (ANL, ARM DQ Office, ORNL) 179 pull requests to add code 4700+ downloads Adopted by ARM in 2020
4. Path Forward § Grow the user base and contributors § Improve code coverage § Want to build up the modules even more, especially retrievals, corrections, and quality control § Updating the roadmap every 6-months https://github.com/ARM-DOE/ACT pip install act-atmos conda install -c conda-forge act-atmos