7.2 Processing and Displaying GOES-16 Imagery with Python

Tuesday, 9 January 2018: 1:45 PM
Room 8 ABC (ACC) (Austin, Texas)
Peter J Pokrandt, Univ. of Wisconsin, Madison, WI

GOES-16 ABI (Advanced Baseline Imager) products began flowing over NOAAPORT in a test mode in March 2017. Since the GOES-16 products were not in McIDAS Area file format or GINI image format, as all earlier generation GOES products had been, the data display and analysis tools in use at UW-Madison AOS (GEMPAK, Unidata AWIPS2, Unidata IDV, etc.) were unable to display these new GOES16 products.

Because the GOES-16 products were in netCDF format, any software that could interpret netCDF files, either natively or via the netCDF libraries should have been able to display the products. However, the products on NOAAPORT were composed of many tiles, subsections of the original image, that needed to be reassembled before plotting. None of the software that we were using to display satellite imagery at the time had this capability built in.

Using python and its interface to the netCDF libraries, with the fast math and array tools contained in numpy, the plotting capabilities in matplotlib, and mapping/georeferencing capabilities in cartopy, we were able to read, reassemble, plot and overlay maps on a product from one of the ABI's visible channels within a few hours of the data first becoming available.

Subsequent code modifications and enhancements have allowed for plotting of the 16 different frequency bands available from the GOES-16 ABI, plotting user defined sub regions from the whole images, displaying with different color enhancements, discovery of, and subsequent fixes for some bugs / data quirks associated with the preoperational aspect of the products, pseudo-color composite imagery, and side-by-side display of multiple bands to more easily understand what they reveal about earth's atmosphere.

Since March, other GOES-16 display and analysis tools have become available, but for background processing (e.g. generating imagery for display and animation in a web browser) we prefer the speed and programatic ease of these python tools.

A description of the data access, plotting and mapping process will be shown, and advantages and limitations to this method of GOES-16 data display and processing will be shared.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner