Tuesday, October 4, 2016

Do people understand hurricane forecast maps?

It seems that we'll see hurricane Matthew landing in the U.S. pretty soon. Mashable has just published an article about how to read a hurricane forecast map. It links to this 2007 NOAA studywhich explains how readers misinterpret it. One of the authors, Kenny Broad, is a colleague of mine at the University of Miami.

NOAA's study focuses on the white cone of uncertainty. Apparently, many people believe that the boundaries of the cone are like the boundaries of the hurricane itself, which is far from the truth:

Both Mashable's article and the study mention another map displaying the probability of experiencing hurricane force surface winds (below). Has anybody tested these maps —and others— in controlled experiments recently? It'd be interesting to see the results. If you know of studies about this, please post them in the comments section. If, as I suggest in The Truthful Art, one of the goals when designing a visualization is to reduce misunderstanding and ambiguity, cases like this are of critical relevance.


  1. I'd love to see more studies about how people read specific common graphics (beyond the generic pies-vs-bars stuff) "out in the wild."

    Amy Griffin and colleagues have been studying how urban planners use Census datasets:
    including some uses of uncertainty on maps. They also studied whether the planners even tend to look at the uncertainty info at all; some admitted they just delete margins of error as soon as they download the dataset.