I am the lead author on this paper. I need to flag an error in reference #35, which is the reference to our own data and software used to produce this article. The URL listed for that reference is wrong. The correct URL for our Zenodo data + software collection is https://doi.org/10.5281/z....

We have informed PLOS of this error, which was introduced accidentally by PLOS in the final stage of the article's production. Their response, however, was that their policies do not allow formal corrections to fix errors of this nature. I feel it is necessary to point this out since we explicitly discuss journal editorial policies in the "Study Limitations" section of our article. Inadvertently, PLOS has emphasized that point for us. Since I am making this comment, I will also point out that PLOS strips DOIs out of all references. They use them to create the "View Article" link below the references, but take them out of the references themselves. This is another example of a journal editorial policy that impacts whether DOIs are listed in references or not.

For a complete and accurate list of all of the references in our paper, please see the article pre-print posted in the NCAR OpenSky institutional repository: http://n2t.net/ark:/85065...