Working on PhD dissertation and for the life of me I cannot figure out why I am getting the error:

Runaway argument?
{NOAA} et~al.(2019{\natexlab {{}}){NOAA}, {ESRL Physical Science Divi\ETC.
! Paragraph ended before \@lbibitem was complete.
<to be read again> 

and another error:

! Argument of \@lbibitem has an extra }.
<inserted text> 
l.496 \bibitem[{NOAA} et~al.(2019{\natexlab{}}}

I checked my bibtex file to see if there are any missing curly braces and there are none. I checked if I have an invisible character and I have none.

I have read on many forums for 24 hours and some said that they may be a package that I have included that will overwrite existing settings. I am not knowledgeable in that field but I referenced the ones of the forums and none of those overwrite what I currently have. Any help with this is appreciated.

As I note, I have obtained a "thesis" document class online and tweaked it to accomplish what I wanted in format.

Below I attach a link to the MWE files:


  • 1
    It is generally preferred not to link to external sites for code examples. Ideally you would reduce your example so far that it fits into the character limit imposed by this site and can be posted directly in the question body. Chances are the issue here is only related to your bibliography style (\bibliographystyle) and an entry in your .bib file (the one containing ESRL Physical Science Divi) and little more, that should easily fit. ...
    – moewe
    Jan 26 '20 at 8:21
  • 1
    ... If you absolutely have to link to external sites, please link to a text-sharing website like pastebin.com or gist.github.com that can be accessed without account and where files can be viewed without downloading.
    – moewe
    Jan 26 '20 at 8:22
  • @Mico and moewe I am sorry for the confusion. I thought the link would work since I have used it before, but apparently it is only restricted to work community. I believe I have successfully uploaded the files to github and you can now view it without downloading it.
    – user966349
    Jan 26 '20 at 8:49
  • @Mico Thank you for the help. I have corrected the errors you mentioned and the } and " that was off was my fault from trying to fix it. How did you find the extra jacox2016impacts? I checked it via a website and that one did not flag it https://biblatex-linter.herokuapp.com. Either way, I fixed the said errors and I still get the same message. I have updated the files that I have linked. I am sorry for being bothersome, but the error and warning messages seem clear. However, I cannot find the runaway argument and the extra } that it claims on line 495. It seems balanced to me
    – user966349
    Jan 26 '20 at 10:10
  • You asked, "How did you find the extra jacox2016impacts?" Simple: BibTeX issued a clear error message about the duplicate entry. Maybe the site https://biblatex-linter.herokuapp.com isn't as thorough as it should be...
    – Mico
    Jan 26 '20 at 10:42

Some comments and observations:

  • The error and warning messages issued by BibTeX should tell you all you need to know in order to fix up the syntax errors in the bib file. E.g., the unbalanced quotation mark in line 20, the unbalanced curly braces in line 494 (hint: replace the second instance of } with "), the repeated entry named jacox2016impacts, the spurious comma after Williams, Claude N. in entry modeldata, etc.

  • There are a lot of issues related to spelling and upper-casing in various bib entries. E.g., the journal field in the entry zou2005regularization currently reads

    journal={Journal of the royal statistical society: series B (statistical methodology)},

    That should definitely be

    journal={Journal of the Royal Statistical Society: Series B (Statistical Methodology)},
  • In the entry OLWR_Equatordata, the field month = "aug", should be month = aug,, i.e., the quotation marks around aug are spurious.

  • Lots of words still need to be encased in curly braces in order to prevent them from being rendered in lowercase due to BibTeX's "sentence style". Some examples: California (several instances), Canadian, El Ni{\~n}o, Southern Oscillation, ENSO, MEI, STL, Pacific, US Pacific Northwest, Southwestern United States, FIO-ESM, etc etc.

    The failure to protect against inappropriate lower-casing, while technically not a syntax error, betrays a worrisome lack of attention to detail. The observation that you probably obtained the entries from some site is not a good, let alone a valid excuse for such sloppiness, to put it bluntly.

  • Quite a few entries (especially those of type @misc) have both a howpublished field and a url field. In all of these cases, the contents of the howpublished field (a) seem to duplicate the contents of the url field and (b) fail to encase the URL string in a \url directive. In such cases, it would seem appropriate to omit the howpublished field.

  • Lots of entries (almost two dozen, I think) have a note field, but the material in these note fields frequently doesn't seem to be appropriate for inclusion in the bibliography. You might open the bib file in an editor and change all instances of Note = with xxNote =. (BibTeX ignores fields whose names it doesn't recognize.)

    At the very least, you should check if the note fields are appropriate. In some cases, the note fields contain ISBN-related information. If that's the case, it's incumbent on you to replace, say, Note = {ISBN 0-387-95457-0} with isbn = {0-387-95457-0}.

  • Out of the 75 entries that contain a month field, fully 70 were published in aug, i.e., August. Is this an amazing coincidence?

  • Last but definitely not least, there's a serious issue (bug??) caused by the fact that you have 42 [!!] entries with the same set of authors, viz., "NOAA, ESRL Physical Science Division, and CIRES" and the same publication year, viz., 2019. I do believe it's a bug, in the sense that the authors of the IEEEtranNS bibliography style and/or of the natbib citation package do not appear to have considered the possibility of a bibliography ever having more than 26 entries that share the exact same author and year fields. In consequence, the year-letter labeling scheme employed by natbib -- "NOAA et al. (2019<X>)", where <X> is a letter between a and z -- runs out of letters once it hits the 27th entry with same author and year field. The resulting error messages (which refer to mysterious (2019{\natexlab {{}}) objects) are rather unintelligible and useless. :-(

    I haven't done enough further digging to be able to tell if the bug is caused by the IEEEtranNS bibliography style or the natbib citation management package. I'm afraid I'm in no position to attempt a true fix of this bug.

    To come up with a short-term workaround, though, do ask yourself if it's absolutely essential for your thesis to list all 42 entries authored by "NOAA, ESRL Physical Science Division, and CIRES" and published in 2019. Do you truly cite (and discuss) all 42 of these entries in the thesis document? Would it maybe be just fine to list roughly half of these entries, or maybe even just a third of them? If it's somehow important to list all 42 entries in question, would it maybe make more sense to provide a separate listing in the body of the document instead of in the bibliography? Just a thought.

  • Thank you for the help. The main issue was the bug with natbib and/or IEEEtranNS.
    – user966349
    Jan 26 '20 at 23:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.