Atmospheric Data Community Toolkit (ACT): A Python Library for Working with Atmospheric Data

The Atmospheric data Community Toolkit (ACT) is a library of codes in the python programming language for working with atmospheric science datasets of varying dimensions. Development efforts in the past have routinely been created in a program or group specific silo space and not with the open source community in mind. This has led to a lot of redundant effort spent on developing code for working with datasets when the effort could be better spent on the science. The goal of ACT is to provide a platform on which the entire atmospheric science community can work together and develop code for applications throughout the entire scientific process, which include

Discovery – Using APIs to find and download data. Currently, the API for the Atmospheric Radiation Measurement Program Data Live Web Service is being used.

I/O – Scripts to ingest data into the common data model. Scripts to read netcdf, csv, and ascii into the common data model have been developed and tested to work on data from a variety of programs.

Quality Control – Scripts for working with bit-packed quality control information and for applying tests to datasets.

Corrections – Scripts for applying corrections to the data. Some instruments require that proper corrections be applied before the data are to be used in a quantitative sense.

Retrievals/Calculations – Scripts for adding value to the data. For example, calculating the stability parameters from a sounding profile.

Visualizations – Scripts for visualizing data in a number of different ways including, 1D and 2D time series, wind rose, histograms, skew-T, and more.

ACT uses a common data model that is built off of xarray and inherits all the functionality of it. The code is self-documenting, using Sphinx to create documentation and unit tests are deployed to ensure that code contributed back to ACT does not break existing functionality. The initial development has focused on reading and visualizing data whereas the upcoming efforts will seek to improve existing capabilities and add for content to the areas for processing and analyzing data (QC, Corrections, Retrievals).

Code Repository: https://github.com/ANL-DIGR/ACT