[ExI] four seasons tree

[Brian Manning Delaney]

Spike,

Very important topic!!

> How do we data-fy birds?  Bugs?

Well, the birds part has some cool data being gathered right now:

http://ebird.org/content/ebird/


eBird is permitting observers to quantify more than species and number 
observed as time goes on. Right now it's "breeding codes" (nest 
building, mating, territorial defense behavior, etc.), sex, approx. age, 
and a few other things. But more options will be added soon.

Sweden has an amazing system called the Species Portal:

http://www.artportalen.se/

It enables people to record data on all life, not just birds.

I became a crazy-avid birdwatcher a year ago, and have used both eBird 
and ArtPortalen, and have spoken with people running both projects. 
There's a problem. This is just an empirical observation. The why of it 
I don't know. With ArtPortalen, people tend to report rarities only. 
"Cool butterfly! I'll log in to my ArtPortalen account and report it!" 
What's needed are systematic observations of ALL butterflies (and not 
just butterflies) in a given area in a given period.

With eBird, people tend to make systematic observations, but only of 
birds, of course. Perhaps it's just because it's less overwhelming to do 
so about birds only, or perhaps the interface is easier. But if we're 
going to ditch the ArtPortalen approached, we'd need to create eBee, 
eButterfly, eTree, etc.


> Perhaps we could get a jillion citizen scientists doing this kind of automated observation.

The more automated, the better. We here tend to be optimists, and some 
reading this might think: "Forget the citizen! Just ask the NSA, GCHQ, 
FRA, all 7-11s, etc. to let us run some recognition algorithms on camera 
footage and the like. Surely such algorithms will be good enough by 2020 
or something."

The people at eBird think high-quality automated recognition of both 
images and sound is well over a decade away. And using images from 
surveillance cameras will be trickier than using images from an observer 
taking a high-quality picture. So we might not have a lot of automation 
soon.

Meanwhile, check out eBird. Lots of fun.

Oh, hey, wait. Maybe image-recognition isn't so far away:

http://ebird.org/content/ebird/news/merlinvision201611/

Pardon typos
Brian