Software for Working with ARM NetCDF Data
Most ARM data are archived and made available as time-series data in netCDF format. NetCDF (designed by Unidata) is relatively compact, appendable, capable of storing descriptive "metadata" along with measurement data, and is platform-independent. Resources for assisting with the reading, writing, displaying, and manipulation of ARM NetCDF data are available on this page.
Learning About netCDF and Obtaining netCDF Libraries
NetCDF software libraries are available at the Unidata website along with information for working with NetCDF data in a wide variety of programming languages:
Justin Monroe, from the ARM Data Quality Office at the University of Oklahoma, led a tutorial on ARM NetCDF files and how to navigate the ARM Python notebook during the 2014 Science Team Meeting.
Getting Started With ARM Data
Once you have a basic toolkit for reading and writing netCDF files, the format is very convenient to work with. To help people get started with ARM netCDF files, we have developed a short tutorial using the Python programming language. This tutorial steps through setting up some basic tools and illustrates how to read a simple ARM data file.
The ARM data usage tutorial uses the Python programming language because it is freely available, but simple-to-install packages are available for low cost.
Giri Palanisamy, Oak Ridge National Laboratory, presented this introduction on how to submit, discover, and acknowledge ARM data, during the 2014 Science Team Meeting.
Preview Plots of ARM Data
To provide the means to rapidly assess the content of ARM data files without using any software, or even downloading data, quick-look plots are generated for many datastreams on a routine basis. These plots give a preview of the file contents. Quick-look data plots are available through the Data Quality Office Plotbrowser tool or through the ARM Data Discovery tool after selecting a set of data to order. Once you data have been ordered from the ARM data archive, the NCVWeb online tool provides the means to perform a variety of data extraction, data conversion, and data visualization tasks.
Open Source Programming
Many scientists are working with ARM data, and it is our desire to share experiences to accelerate the collective application of these data. A good way to support this community development is to make a practice of sharing codes, especially for routine processing activities that many others might need.
An example of an OpenSource code suite is the Python ARM Radar Toolkit (Py-ART). Py-ART is a toolkit that combines a variety of utilities for processing ARM radar data. These utilities are available through an ARM account on GitHUB.
Py-ART is the first ARM open source project, but we hope others will engage in community software development and invite others to make use of the ARM GitHUB account.
Scott Collis, Argonne National Lab, led the first-ever Python ARM Radar Toolkit (PyART) short course during the 2014 Science Team Meeting. This course covers an intro to radar data and what it looks like in PyART, plotting RHIs and basic analysis, and an at-home exercise.
The ARM Archive Data Visualization Cluster
In part to support community software development, but particularly to provide a mechanism to work with large volumes of data without having to download them, a public software development space has been set up at the ARM Data Archive. This Data Visualization Cluster provides programming ability in various languages including Python, Matlab, and IDL, as well as specialized software including the Py-ART radar toolkit.
If You Still Have Questions
If you still have questions about using ARM data, there are many individuals associated with ARM and the Atmospheric System Research (ASR) program who have extensive experience with these data. A good place to start is to contact the technical lead (also known as instrument mentors) for the instrument you are interested in or one of the data translators who serve as liaisons between the science community and ARM data product software developers. You can also send a question to email@example.com and we will try to connect you with the right person.