Open Thread

Here is some new material based on the Band 13 and Band 16 radiance data files for the CONUS region from the GOES East geostationary satellite.  Each archived data file is a 2500 column by 1500 row array of numerical values in units of mW m-2 sr-1 (cm-1)-1 (in words, this means milliwatts per square meter per steradian per wavenumber.)  

Band 13 is the “10.3 micron Clean Longwave IR Window” channel. 2 km resolution, 5-minute update. 

Band 16 is the “13.3 micron CO2 Longwave – IR” channel. 2 km resolution, 5-minute update. 

For this exercise, the first of the 12 data files in each 1-hour block of archived files was downloaded and processed into a series of images and histograms. There are 168 files at these 1-hour sampling intervals running from 7-1-2026 at 0000Z to 7-7-2026 at 2300Z. 7 full days of summer conditions. The images and histograms from the Band 13 and Band 16 data are merged into a set of 168 sequential side-by-side pairs for ease of time-matched comparison. The files are named numerically in time series order.

In the folder below, there are subfolders with the images and histograms. In the main folder, there is the custom color scale and two plots. One plot gives the mean radiance for the entire region, by hour, for Band 13 and for Band 16. The other plot expands the vertical scale and shows the mean, the 5% quantile, and the 95% quantile. (Special note: In this same folder, I have also placed a plot of “radiance v wavelength…” from MODTRAN. It uses magenta shading for the Band 16 range of wavelengths, for which a significant portion of the static “warming” tendency of the 2XCO2 case is computed. Readers here at WUWT might remember this plot from a few years ago.)

https://drive.google.com/drive/folders/11jRax2RWdZe90Sqf84NGt_Q0wMzfdHnI?usp=sharing

Source data files are from:

https://noaa-goes19.s3.amazonaws.com/index.html#ABI-L1b-RadC/2026/182/

The purpose was to directly compare the longwave emitter performance and characteristics for these two bands, using time-matched data files. That is why the same custom color scale is used for both bands for the images.  

Discussion
Stepping through the images, the influence of active cloud formation and dissipation on IR emission from any part of the region is obvious. Also obvious is the rapid rise and decay of IR emission to space from under clear skies as the daytime pulse of solar energy is absorbed at the skin surface, followed by rapid cooling. This is more pronounced in the window band, as one would expect.  Not so obvious is that the radiance above the cloud tops gives a higher value (i.e. a bit less “white” in the images) in Band 16 than in Band 13.  

The region-wide mean for Band 16 is greater for all hours than for Band 13. Interesting. Note also the differences in the 5% and 95% quantiles. I don’t yet have a well-formed explanation for these findings. 
 
So without further comment for now, I am posting this material for those who are interested in how the emitter is observed to perform. 
 
NOAA describes the 16 bands for which data is acquired from the satellite at the following web page.
https://www.star.nesdis.noaa.gov/GOES/conus.php?sat=G19&refresh=true

Thank you for your interest in this matter.

P.S. Grok assisted by generating a list of valid URLs for automated download of the specified files. Doing so manually would take many hours, so it deterred this exercise until now. Grok also assisted by suggesting code snippets in R for image generation and for merging, and for generating and applying the specified color scale. This is where AI can be very useful.