search

SimonGoring / WitnessTrees

Key working directories for the initial data products of the PalEON Project.
http://paleonproject.org
MIT License
0 stars 4 forks source link

no tree questions #2

Open paciorek opened 9 years ago

paciorek commented 9 years ago

1) Simon, I'm not sure what you mean by the boundary in no tree between S MI and the rest of the state. In the attached image (which shows proportion of "no tree" in each grid cell) I don't see anything.

2) In attached image, I'm concerned about the red spot in the UP. Can we explain this?

3) In the attached image, I'm concerned by the lack of "no tree" in the area around Chicago. We got those data from another source - could the lack of "no tree" relate to that?

SimonGoring commented 9 years ago

@paciorek there are no attached images.

paciorek commented 9 years ago

I can select my pdf but it doesn't seem to be attaching it. I'm emailing it to you.

paciorek commented 9 years ago

Simon says: When it's gridded you can't see it as well, and I suspect it's actually hidden in the lowest (dark blue) class, can you add a second color for '0'? There are a bunch of isolated "No tree" plots in the Lower Peninsula, but as soon as we cross into the new 'Southern Michigan" domain they disappear almost completely.

Chris: now I see it, when I just plot 0 no.tree vs any no.tree. It's also very stark at the WI-IL border where we go into the Chicago data that have no no.tree.

SimonGoring commented 9 years ago

Yes. Thanks Chris.

I think it's something that we'll need to (1) talk about face to face, and (2) develop some sort of test to check to see how critical the issue is.

On Thu, Dec 11, 2014 at 3:57 PM, Christopher Paciorek < notifications@github.com> wrote:

Simon says: When it's gridded you can't see it as well, and I suspect it's actually hidden in the lowest (dark blue) class, can you add a second color for '0'? There are a bunch of isolated "No tree" plots in the Lower Peninsula, but as soon as we cross into the new 'Southern Michigan" domain they disappear almost completely.

Chris: now I see it, when I just plot 0 no.tree vs any no.tree. It's also very stark at the WI-IL border where we go into the Chicago data that have no no.tree.

— Reply to this email directly or view it on GitHub https://github.com/SimonGoring/WitnessTrees/issues/2#issuecomment-66696495 .

paciorek commented 9 years ago

ok, let's chat this weekend. I would like to understand how it's possible that so many survey points in praire such as near Chicago could possibly all have two trees.

Here's a grid cell near Chicago that has only 17 trees. x y alder ash bald.cypress basswood beech birch black.gum black.gum.sweet.gum buckeye cedar.juniper cherry chestnut dogwood elm fir hackberry hemlock hickory 15188 605000 598000 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ironwood locust maple mulberry no.tree oak other.hardwood pine poplar poplar.tulip.poplar spruce sweet.gum sycamore tamarack tulip.poplar unknown.tree walnut willow 15188 0 0 0 0 0 16 0 0 0 0 0 0 0 0 0 0 0 0

There are other points like this. I'm guessing the Chicago data don't have survey points being recorded in the raw data if there were no trees. For composition this is not a big deal but obviously for the other outcomes it's huge.

Who's the manager of the Chicago data? Jody?