Very recently, I got the following note from a student. While such e-mails arrive on occasion, this one (in bold) was uniquely thoughtful and thought-provoking, as you’ll see in my somewhat long-winded response, reproduced below it in italics. This represents a strong concern I’ve heard from several undergrad and grad students over this decade, and it’s likely just smoke from a larger fire of worry out there in student-land. So I’m addressing it here for a broader audience, using that correspondence.

This entire package has become a second supplementary BLOG entry here to the original post (read first). [Here was the first supplemental entry, from 2018 (read second).]

I’m writing you as a junior mathematics student who has (for most of my life) had an absolute fascination with weather, particularly severe convection, and the forecast process. I enjoy forecasting severe weather as a hobby and pastime and am considering graduate school in meteorology. My question is: What do you view as the future of the role of humans in the forecast process? What I’ve gathered from several sources, including your blog (which I enjoy reading a lot, by the way), is that humans will likely always play a role in the process, especially in short term and high impact forecasts. Do you think this is an accurate view, especially in light of initiatives like NWS Evolve? I would hate to pursue this field professionally only to find out in 5–10 years that the job I want doesn’t exist, or doesn’t exist in the form that I would want to work in.

Thanks for writing.  My personal thoughts outlined on that BLOG entry (which of course don’t necessarily represent my employer) are still valid in the era of “NWS Evolve“.  The direction clearly is more toward automating single-variable forecasts like temperature, wind, and so forth, especially days out, and converting more and more forecaster time to short-fused, targeted, higher-impact priorities like warnings and watches, and to what’s called “Decision Support Services”.  That’s basically bureaucratic lingo for communicating with power users of forecasts:  emergency managers (local, state, FEMA), media, law enforcement, other government agencies at all levels, and direct-to-public engagement (such as social media and online briefings).

Understanding of meteorology still will be as important as ever.  It has to be.  You cannot fully and effectively communicate what you don’t understand.  Some people with highly polished speaking or writing skills, but deficient in understanding, can skate by for awhile with audiences of less expertise.  They look and sound like they know more than they do.  However, eventually they will make a mistake based on that lack of knowledge, and lose credibility for preventable reasons having nothing to do with normal forecast error.  In what we do, credibility is everything.  Do everything you can to earn and keep it!

That’s why your undergrad math and graduate meteorology degree(s) won’t be useless, but MUST be supplemented with good communications skills—written and verbal.  I can tell from the quality of your e-mail that you’re off to a good start on the written side.  Keep it up!  Enhancement of writing skills (even if already good, like yours) is a necessary and career-long process.  Take advantage of opportunities for public speaking and accepting critiques of that, too. The higher-impact the forecast, the more communications skills matter. 

I know it’s an overused cliche, but it’s true in this science:  it also helps nowadays to learn to code.  Back in the late ’80s and early ’90s when the NWS forecaster depended more completely on applications of science theory, I didn’t have to.  You will.  Fortunately any good degree program will include that as part of the curriculum, and I hope you’re already taking advantage of both required and elective programming opportunities.

Communication also is increasingly important on the forecast floor, as the process is more cooperative and multi-person in nature than ever.  [I’m trying to avoid the bureaucratic buzzword, “collaborative”.  :-)]  I lean heavily on my colleagues with good coding skills in my research and forecasting, through the tools they develop.  I depend on them in live situations, for specific meteorological experiences I still haven’t seen.  In turn they learn from my decades of experience, conceptual understanding, attention to analytic detail, and deep immersion in the science.  Anymore, forecasting on an island on shift is a dying concept.  Nobody can know everything, and consistent excellence depends on input of others who also have expertise.

I cannot promise the job (especially at a local-office level) won’t evolve out of a form that you would want to work in.  Maybe it will.  Maybe it won’t.  Science conceivably could be de-prioritized or centralized, orphaning the local level to one of weather communicator only, or consolidating scientifically based forecasting in fewer and fewer physical places.  If that happens, I don’t know how long it would take—it’s largely at the whim of higher-level managerial policy and priorities. 

That’s why you need a good backup plan (that I gambled on not having!) to operational meteorology, just in case.  The good news is:  that’s still years off, if at all, and you can and should accumulate the skills you’ll need to adapt to that and any other reasonable contingency.  But don’t abandon the dream before pursuing it, just because the job might change in unfulfilling ways.  It may not, or you could find yourself in an unforeseen opportunity that turns out to be a “blessing in disguise”.  In the meantime, I find no better calling than providing the best forecasts humanly possible to the local offices, media, storm spotters, and EMs, and through them, the taxpayers at large who depend upon us.  If you are thinking along those lines, then please, go for it, all-out.  We and the taxpayers certainly could use such talent and motivation.

Sorry for the verbosity…you got me going on a near-and-dear topic. 
🙂

I hope this helps. Good luck with your studies!

===== Roger =====