[obol] Re: Thrush migration maps "cool" but way too smooth
- From: David Irons <llsdirons@xxxxxxx>
- To: OBOL Oregon Birders Online <obol@xxxxxxxxxxxxx>
- Date: Wed, 30 Apr 2014 08:51:09 +0000
Greetings All,
Joel has offered some good cautionary points about animation maps. The issues
associated with producing these sorts of renderings are not limited to
animations that show bird migration routes. Invariably, the datasets that are
used to create visual representations for all sorts geographic phenomena are
not nearly as robust as you might think. It's hard to imagine how many years
might pass before the avian datasets will be robust enough to produce these
sorts of animations without the need for algorithmic smoothing. It probably
won't happen in my lifetime.
As Joel points out, you cannot use these animations as a baseline for
evaluating the validity of individual reports. Similarly, you cannot use the
range map in your field guide to predict whether you might find a species at a
specific point within the broader range for that species. I presume that this
is understood by most of the folks in this forum.
In simple terms, a range map or one of these animations is much like a slice of
Swiss cheese, there are lots of holes. You might equate the holes to an absence
of positive data, or in many cases the absence of any exploration whatsoever.
Birders and bird researchers tend to not sample the landscape evenly, hence the
"clusters" of observations Joel describes. Continuing the Swiss cheese analogy,
one way to think about the smoothing algorithms is that they are like applying
heat to the slice of cheese. The holes fill in as the cheese melts and
redistributes more evenly across the surface it's on. This smoothing process
results in a more attractive map/animation that offers a reasonably accurate
macro view of a pattern, but fails to accurately depict what is happening
within every latilong inside the map area.
Since we humans are visual learners, these maps can be helpful in providing us
with at least a rudimentary understanding of the geospatial aspect of migration
routes. They may also, to a lesser extent, help us appreciate when birds reach
certain latitudes. It takes many hundreds, if not thousands of words to
otherwise describe the timing and route of a migration. On this level, these
animations have value, despite their inherent flaws. If you want to truly
understand what is happening at any given point within these maps, you will
need to dig much deeper into the data (presuming that it exists for the point
that you are interested in), or better yet, go explore for yourself. If you
endeavor to go birding where others don't, you'll almost certainly make
observations that accurately fill some of the holes in the Swiss cheese.
Dave IronsPortland, OR
Other related posts:
