[obol] Re: Mining eBird data
- From: Priscilla Sokolowski <priscillanhk@xxxxxxxxx>
- To: Mark Nikas <elepaio@xxxxxxxxx>, OBOL <obol@xxxxxxxxxxxxx>
- Date: Sat, 21 Mar 2015 07:51:28 -0700
Mark and others interested in this,
Someone mentioned the BIRDSEYE app, which I have and enjoy. But just as
with birdtrax,
none of the apps I have seen (three or four), will go back in time further
than 4 weeks.
(Maybe I'm wrong about this, but I don't think so. Most apps only go back
as far as 4 weeks because
of some limitation in what app programmers can get from the eBird database
as it is currently set up.
I don't think any apps provide a fully historically based bird sighting
list for small regions).
Looks like at this point the method you have used and the method Vjera
showed you
(which I am not familiar with) seem to be the only way to collect
historical data for a small region
which is not an eBird IBA. (I don't know how to do what Brandon was
suggesting to find the checklists
which underlie an IBA bar chart in eBird but I posted a question about
that).
GPS is a mixed blessing. Several of the birding apps will grab gps
coordinates when you record a bird sighting
(or on some you can edit the map sighting location for a bird to get the
gps coords). While gps does give
a precise location of a sighting, that very precision makes it difficult to
associate with a larger region,
until someone does some coding which can do exactly that. I guess that's
what Brandon's talking about with eBird's IBA concept. This makes me think
of the CBC count circle app, which can show you whether or not your current
location is inside or outside of a particular CBC count circle. I don't
think the coding for making that discrimination is all that complicated,
but what is complicated I'm sure, would be to integrate that kind of coding
into an existing program like eBird. The contrast between WYSIWYG website
design methods vs direct html website design comes to mind as a metaphor
for this situation. We don't seem to have any kind of WYSIWYG sort of way
of processing bird sighting data. You know, "click here and drag a circle
of desired size" and the program collects all data which lies in that
circle. Sounds so simple doesn't it? Boy that would really be neat! Some
kind of analog (regional, continuous) solution to dealing with digital
(discrete) bird-sighting data.
I like the suggestion (made by Mike Patterson I think), of using
birdnotes.net. Especially the idea of creating some kind of region of
interest for the Diamond Hill area. Seems like that could be a great tool,
once other birders contributed data too. Joel Geier has often mentioned
that site, but because he usually did so in the context of discussing the
history of eBird, I did not realize that birdnotes was still actively used
by many fine birders in Oregon. When I went to the site a year ago I was
surprised to see how many birders were using it. (I had thought it was
"extinct", and no longer used). To me, birdnotes is the "best-kept-secret"
and should be given more publicity. I cannot say this without also saying
the same thing for the site in which Dave Irons and others post those
highly informative articles on bird identification as well as photographs
depicting the important details. What is the name of that site ...
BIRDFELLOW.COM. I have not made nearly as much use of that site as I
should.
With the birdnotes site, I have "mined" all the short-eared owl historic
data I could find for the Diamond Hill Wetlands area and found a handful of
significant sightings/reports which were not in the eBird database.
In particular, sightings outside of the "wintering" season, which could
suggest the possibility of breeding.
(And even if not, that site is a new wetland and the habitat is still only
getting better for that possibility in the future).
Mark, it seems your original question really hits on an important (and
growing?) problem with existing databases and birding apps. This isn't
going to go away until solutions begin to emerge.
Priscilla Sokolowski
On Fri, Mar 20, 2015 at 11:52 AM, Mark Nikas <elepaio@xxxxxxxxx> wrote:
> Thanks for reminding me of BirdTrax Priscilla. I used to use iGoogle as
> my computer home page and I had several BirdTrax widgets on it to track
> local birding sites of interest. iGoogle went away a couple years ago and
> I never found a suitable replacement and I forgot about BirdTrax. I'll have
> to set it up again. I can figure out how to set up a simple web page to
> use it on.
>
>
> Mark
>
>
>
>
> On Fri, Mar 20, 2015 at 8:01 AM, Priscilla Sokolowski <
> priscillanhk@xxxxxxxxx> wrote:
>
>> Mark;
>>
>> There is a way to set a radius circle centered at GPS coords of your
>> choice
>> and a time span over which to look, using BIRDTRAX.
>> This is not an "app" but some code one puts on a web page
>> after setting the parameters one wishes; center of circle, number of days
>> to look back; and a few other things.
>> You can view the page with the code running on it using any device with
>> an internet browser.
>>
>> What I especially like about it is that unlike most searches within eBird,
>> this one reaches across county lines.
>>
>> I have a circle set up to cover a 30 mile radius which encloses all the
>> areas I
>> would readily drive to. You can see it on my "updates" page.
>> Scroll down to see the birdtrax thing. Then scroll down in the birdtrax
>> box
>> to look back through time (up to a week the way mine is set up).
>>
>> Take a lot at the one I have set up and see if it does the kinds of
>> things you want.
>> If you like it and don't have a webpage you can set one up on,
>> I will gladly configure one as you want it and host it on a separate
>> page and give you the link to it. Any time you want a change
>> in the settings I can easily and quickly make those for you.
>>
>> If you click on the "+" sign at the far right of a bird
>> you get the options to go to the ebird (website) checklist or map
>> for that particular sighting.
>>
>> If you click on a species you will get a sublist of all reports of that
>> bird
>> in your circle in the time period you have set.
>>
>>
>> Here is where I got it:
>> http://www.birdventurebirding.com/p/birdtrax.html
>>
>> Priscilla Sokolowski
>>
>>
>>
>> On Thu, Mar 19, 2015 at 8:05 PM, Mark Nikas <elepaio@xxxxxxxxx> wrote:
>>
>>> Thanks all for your responses. I thought I might be missing something
>>> obvious but apparently not. Interesting that an external app can mine eBird
>>> data better than eBird itself. At least for the folks who enter the data in
>>> the first place. One more thing to nudge me closer to getting a smart
>>> phone. I'm pushing back though.
>>>
>>> I have used the brute force method of searching for common species to
>>> find the checklists entered for a specific area. That's how I came up with
>>> all the pseudonyms for Diamond Hill in the first place. The increasing use
>>> of coordinates over place names is making this method ever more time
>>> consuming though.
>>>
>>> When it was just place names being used the different place names could
>>> be merged in a search. When requesting a checklist for Fern Ridge Res. for
>>> example you could select any or all of the several different hot spots
>>> scattered around the Fern Ridge area to come up with a more useful list for
>>> the area (thanks Vjera for showing me how to do that). This method excludes
>>> data entered under coordinates though.
>>>
>>> Mark Nikas
>>>
>>> On Thu, Mar 19, 2015 at 7:35 PM, Mike Patterson <celata@xxxxxxxxxxxx>
>>> wrote:
>>>
>>>> One of the as yet to be solved issues regarding eBird data is its
>>>> lack of easily extractable data at resolutions finer than county.
>>>> Oregon counties are large and unwieldy unlike East Coast counties
>>>> where most birders (and apparently most eBird programers) live.
>>>> Several of us have brought this up (once or twice) in previous
>>>> eBird discussions....
>>>>
>>>> I have mostly used brute-force methods to extract data. With Bean Goose
>>>> data, I pulled up the map in "Explore Data > Species maps" and then
>>>> clicked on every balloon and wrote the info I wanted on a legal pad. I
>>>> then put it into a spreadsheet myself.
>>>>
>>>> You can probably get a close approximation species list by picking
>>>> a common species (like Song Sparrow) for a species map and then zooming
>>>> in to the Diamond Hill area. You'll still have to hand capture all
>>>> the lists and you may need to pick more than one common species to get
>>>> enough lists.
>>>>
>>>> You could also go to birdnotes.net and generate a checklist for Diamond
>>>> Hill wetlands. But that looks like its mostly populated with data you
>>>> put there, so it may not add anything...
>>>>
>>>> I am told that eBird will eventually have aggregate capacity a finer
>>>> scales using polygons, but don't expect it any time in the next couple
>>>> weeks...
>>>>
>>>> --
>>>> Mike Patterson
>>>> Astoria, OR
>>>> the CODE
>>>> http://www.surfbirds.com/community-blogs/northcoastdiaries/?p=2702
>>>>
>>>>
>>>>
>>>> OBOL archives: www.freelists.org/archive/obol
>>>> Manage your account or unsubscribe: //www.freelists.org/list/obol
>>>> Contact moderators: obol-moderators@xxxxxxxxxxxxx
>>>>
>>>>
>>>>
>>>
>>
>
Other related posts:
