8 years ago: Don’t need a weather man

November 3, 2005, here on slacktivist: Don’t need a weather man

This is almost precisely the opposite of what newspapers exist to do. They exist to provide their readers access to information — not to wedge themselves in between the readers and that information.

But by interfering with their readers’ access to the weather, newspapers are able to keep them on site. Directing readers to the repackaged page means more hits, which means more ad revenue. And ads can be sold on the weather page too.

I appreciate that newspapers have to pay the bills. Directing readers to the NWS site might be a better, and more honest, service, but it results in no apparent revenue. So it might seem that what I am suggesting sounds nice in principle, but is unsustainable as a business plan.

But in the long run, newspapers cannot afford not to do this. Ensuring that the paper can be relied on to provide the most direct access to the best available information earns the readers’ trust. And it is only that trust that will keep them coming back.

  • Ursula L

    Hmm… while I’m not sure what the newspaper in Rochester does with the weather, I know that at least one television station had excellent weather forecasts, for years even better than the national ones for the area. They had their own meteorologists, their own local radar before the NWS set up weather radar in Rochester, (as one of their ads said when they first set up the radar “Don’t be Buffaloed, without Rochester radar you’re snowed…” because the Buffalo radar couldn’t pick up lake effect hitting Rochester from Lake Ontario.) and the forecast included discussion of how fronts were moving through the area, the way the different lakes were interacting to affect the weather, etc. (I don’t know if it is still this way. And I know not every town and city had this quality – it is something I missed whenever we were traveling.)

    The NWS site is very good at telling you what the forecast is. It isn’t as good at telling you why they’re drawing the conclusions that they are drawing, or how, if different variables change, the forecast will change.

    That’s really where the opening is for local news. Not to repackage the NWS forecast, but to explain it so that rather than it being a seemingly arbitrary prophecy of what the weather will be for the next week, the weather is a phenomenon that the average viewer or reader can understand.

    Which in turn is useful in a wider context for society, because it is way in which ordinary people can learn to understand how science works, and see it affect their daily life. Simply relying on the finished forecast doesn’t really give people a chance to understand how data becomes a forecast. If you have a notice that the high will be X with Y% chance of snow, you don’t know that a cold front is coming from the northwest, and it will pick up moisture as it crosses the lake, making it both colder when that happens and bringing snow in the distinctive pattern of lake effect. But if you know how the weather works, it is easier to go from that to accepting more complex results, such as the way global warming works, because you know that there is a method to the apparent madness of percent chances of precipitation and expected temperature.

  • MaryKaye

    Nate Silver writes about this in _The Signal and the Noise_. He notes that few places just reprint the NWS numbers unchanged; there are usually modifications. As Ursula says, some might be well-founded use of local expertise. But SIlver shows that overall, statistically, the further you go from the NWS the more inaccurate you are. As I recall, almost all repackagers bump the NWS’s 100% and 0% up or down a little as they are afraid of criticism if they say something is 100% and then it doesn’t happen. This costs only a little accuracy. But TV stations tend to modify a lot more, and they are a lot less accurate overall. Presumably false weather sells better than real weather. Not much of a public service, though.

  • Ursula L

    The situation I was talking about was a while back – I think that the initial introduction of local radar was in the 1980s. People knew that local radar made for better forecasts, but the NWS hadn’t yet built a full national system with radar stations everywhere. A few years later, the same station built its own Doppler radar, when it was established as helping improve weather reports but the NWS hadn’t yet installed it in Rochester. But the careful way the station presented the weather forecast continued for as long as I was in Rochester – into the early 2000s. They clearly made a priority of hiring good meteorologists, and giving them good tools to work with. And they “good meteorologist” meant not just presenting accurate weather forecasts, but presenting the weather in a way that was accessible and inherently educational.

    My point is not that the local forecast will necessarily be better, but that a skilled meteorologist can present the information that is behind the NWS forecast in a way that educates the public about how weather works, and provides a general understanding of the way in which scientists can draw conclusions and predictions from observed phenomena.

  • rrhersh

    The most interesting bit from the post of eight years ago was that Santorum was corrupt enough to introduce an explicit “fuck you, America!” bill at the behest of Accuweather. Fortunately buying off the rest of Congress seems to have been beyond Accuweather’s means. I will think of this the next time Santorum moralizes at us. I will also think of santorum, of course…

  • Abel Undercity

    I just checked the NWS forecast for my area just to assure myself that Santorum’s bill was a failure.