Error in plotting biophysical values

[teixeirak]

This is on the new version (new plots). Two issues: 1- for some native ecosystem types (have seen it for tropical and temperate forests, and it appears in the attached image), the Rnet term plots as positive. It should be negative. This error might have been introduced when the sign convention on the .csv file was switched. 2- I'm not seeing the biophysical on the third plot. It also is unclear whether the values are adding up correctly. For example, for soy in the attached image, the two negatives should add to give a larger negative, but the black dot indicating their sum is near zero.

[teixeirak]

David, this entry summarizes the problem plotting biophysical values as I understand it so far. I will work on adding more info.

[teixeirak]

This is a plot of typical, average values for a range of biomes:

This is a plot characterizing the range of spatial variability. Based on this, I do not think it is true that it is impossible to have negative Rnet values for agroecosystems (one of the issues Evan raised). However, I'm not sure it is correct, either.

[teixeirak]

Here is a shot of what the output looks like for several of those biomes.

[teixeirak]

ERROR IDENTIFIED!!! The biophysical plot is showing the Biogeochemical values.

[teixeirak]

For the third plot...

My guess is that, for the agroecosystems, the biogeochemical values are being subtracted from themselves to get the net value, always yielding zero.

Something else is going on with the native ecosystem types; these look like they could be correct (forests look reasonable; grassland and desert are different from what's typical, but high spatial variability means that it could be correct).

[dlebauer]

I fixed the error of biophysical plots showing biogeochemical values.

For further fixes, it will be important for me to know if the underlying data are correct.

Also, if you have LE and Rnet for Switchgrass and MxG I can put those in.

If there are missing biophysical values, should I exclude the bar on the third (CRV) plot?

[teixeirak]

David,

There is no one default biophysical value; they are highly variable, and are always drawn from the Snyder IBIS map. Means and standard deviations are shown in the plot that I included in the earlier post here.

There is one way in which those data could be off. These values are all the difference between modeled vegetation and modeled bare ground. In theory, the modeled bare ground should all be the same, but I'm not sure if it is exactly. There is a global version from IBIS (definitely in the calculator) and regional modeled values from AgroIBIS that pair with the agroecosystem model runs. I am not sure if the latter have been added to the calculator. I think I asked Carl to do so, but... ???

I'll look now and evaluate the data.

K

On Feb 27, 2014, at 9:49 PM, David LeBauer wrote:

I fixed the error of biophysical plots showing biogeochemical values.

For further fixes, it will be important for me to know if the underlying data are correct.

Also, if you have LE and Rnet for Switchgrass and MxG I can put those in.

If there are missing biophysical values, should I exclude the bar on the third (CRV) plot?

— Reply to this email directly or view it on GitHubhttps://urldefense.proofpoint.com/v1/url?u=https://github.com/ebimodeling/ghgvc/issues/53%23issuecomment-36317178&k=diZKtJPqj4jWksRIF4bjkw%3D%3D%0A&r=nNtDYeFrH59SuS8xeCmQgA%3D%3D%0A&m=LnanrC3obdeSQDbRgKdtUqFA4CflBt7eegsUg%2BHObA8%3D%0A&s=d4b7e1ed3d2911346f3dd3a589a040cc089d9511f3e849eb47456668f6b5ca36.

Kristina J. Anderson-Teixeira Ecologist, Leader of ForestGEO/CTFS Ecosystems & Climate Initiative Smithsonian Institution Smithsonian Conservation Biology Institute 1500 Remount Rd.; Front Royal, VA 22630 USA 1-540-635-6546 | teixeirak@si.edumailto:teixeirak@si.edu

[teixeirak]

Yes, if there are missing biophysical values, please exclude the bar on the third (CRV) plot.

On Feb 27, 2014, at 9:49 PM, David LeBauer wrote:

I fixed the error of biophysical plots showing biogeochemical values.

For further fixes, it will be important for me to know if the underlying data are correct.

Also, if you have LE and Rnet for Switchgrass and MxG I can put those in.

If there are missing biophysical values, should I exclude the bar on the third (CRV) plot?

— Reply to this email directly or view it on GitHubhttps://urldefense.proofpoint.com/v1/url?u=https://github.com/ebimodeling/ghgvc/issues/53%23issuecomment-36317178&k=diZKtJPqj4jWksRIF4bjkw%3D%3D%0A&r=nNtDYeFrH59SuS8xeCmQgA%3D%3D%0A&m=LnanrC3obdeSQDbRgKdtUqFA4CflBt7eegsUg%2BHObA8%3D%0A&s=d4b7e1ed3d2911346f3dd3a589a040cc089d9511f3e849eb47456668f6b5ca36.

Kristina J. Anderson-Teixeira Ecologist, Leader of ForestGEO/CTFS Ecosystems & Climate Initiative Smithsonian Institution Smithsonian Conservation Biology Institute 1500 Remount Rd.; Front Royal, VA 22630 USA 1-540-635-6546 | teixeirak@si.edumailto:teixeirak@si.edu

[dlebauer]

I understand that Rnet and LE are spatially variable ... but isn't the same true for GHG flux? For tomorrow, would be nice to have something for the plots

[dlebauer]

We have the data but as explained in gh-27 the code Carl wrote to use it is a mess and needs overhaul.

[teixeirak]

It is true of GHG flux, and we have underlying maps (Saatchi tropical biomass in year 2000, global soil C). However, we cannot have these come up as the defaults until we get a good map of land cover. At present, the calculator doesn't know if the Saatchi biomass values (for example) should be assigned to a natural or agricultural ecosystem at a given location.

K

On Feb 27, 2014, at 10:06 PM, David LeBauer wrote:

I understand that Rnet and LE are spatially variable ... but isn't the same true for GHG flux? For tomorrow, would be nice to have something for the plots

— Reply to this email directly or view it on GitHubhttps://urldefense.proofpoint.com/v1/url?u=https://github.com/ebimodeling/ghgvc/issues/53%23issuecomment-36317917&k=diZKtJPqj4jWksRIF4bjkw%3D%3D%0A&r=nNtDYeFrH59SuS8xeCmQgA%3D%3D%0A&m=sMxSSFH1JZ1fvQQoQPodFuixhaJaTbO8m03h5B%2B0eE0%3D%0A&s=7fd86a84c39453fac547544fd02da633227d79499e676efae0bfa9d2867a4134.

Kristina J. Anderson-Teixeira Ecologist, Leader of ForestGEO/CTFS Ecosystems & Climate Initiative Smithsonian Institution Smithsonian Conservation Biology Institute 1500 Remount Rd.; Front Royal, VA 22630 USA 1-540-635-6546 | teixeirak@si.edumailto:teixeirak@si.edu

[teixeirak]

Oh, you mean we should put in average values for the biophysics for the crops? Got it. Good idea-- they'll never know the difference, and it at least gives us a good approximation.

On Feb 27, 2014, at 10:08 PM, David LeBauer wrote:

We have the data but as explained in gh-27 the code Carl wrote to use it is a mess and needs overhaul.

— Reply to this email directly or view it on GitHubhttps://urldefense.proofpoint.com/v1/url?u=https://github.com/ebimodeling/ghgvc/issues/53%23issuecomment-36317969&k=diZKtJPqj4jWksRIF4bjkw%3D%3D%0A&r=nNtDYeFrH59SuS8xeCmQgA%3D%3D%0A&m=Y%2B%2BBeP8AXJPMdfjNL2NCPUbKeikNPi0o6PuPYXzgNsI%3D%0A&s=6c699692e9d32ff13af14c16692d0dfec24898b3f114fce82df57bf7c704c73d.

Kristina J. Anderson-Teixeira Ecologist, Leader of ForestGEO/CTFS Ecosystems & Climate Initiative Smithsonian Institution Smithsonian Conservation Biology Institute 1500 Remount Rd.; Front Royal, VA 22630 USA 1-540-635-6546 | teixeirak@si.edumailto:teixeirak@si.edu

[teixeirak]

Just a minute. I have to dig back in my archives to get that data.

On Feb 27, 2014, at 10:08 PM, David LeBauer wrote:

We have the data but as explained in gh-27 the code Carl wrote to use it is a mess and needs overhaul.

— Reply to this email directly or view it on GitHubhttps://urldefense.proofpoint.com/v1/url?u=https://github.com/ebimodeling/ghgvc/issues/53%23issuecomment-36317969&k=diZKtJPqj4jWksRIF4bjkw%3D%3D%0A&r=nNtDYeFrH59SuS8xeCmQgA%3D%3D%0A&m=Y%2B%2BBeP8AXJPMdfjNL2NCPUbKeikNPi0o6PuPYXzgNsI%3D%0A&s=6c699692e9d32ff13af14c16692d0dfec24898b3f114fce82df57bf7c704c73d.

Kristina J. Anderson-Teixeira Ecologist, Leader of ForestGEO/CTFS Ecosystems & Climate Initiative Smithsonian Institution Smithsonian Conservation Biology Institute 1500 Remount Rd.; Front Royal, VA 22630 USA 1-540-635-6546 | teixeirak@si.edumailto:teixeirak@si.edu

[teixeirak]

Just to make sure its clear, the values you see on this screen shot correspond to the ones that are highlighed yellow in the spreadhseet I just sent you.