Adding global gridded yield data -> updated AgProdChange xml's

[abigailsnyder]

An early iteration of this code existed in the old data system, but was not very flexible and was not translated.

For a couple of my other projects, I've written slightly more flexible code to take global gridded yield data, turn it into agricultural yield multipliers, and apply those to the existing AgProdChange assumptions. This allows us to consider the full cross of RCPxSSP space (RCP via global gridded yields, SSPs via different AgProdChange assumptions).

I think, in general, figuring out a way to incorporate this into gcamdata could be widely useful. At the very least, it would save others from repeating work, and it could help us have a more consistent methodology across projects that do these kinds of experiments.

One of my next projects is possibly doing a GCAM-hector-yield emulation loop. I'm not 100% sure yet (software design + engineering is not my strong suit) but I do think it could be useful to get this processing code into gcamdata as part of that.

[abigailsnyder]

an example R project of this process lives on pic: /pic/projects/GCAM/Abigail/isimip_pdssat_impacts

[abigailsnyder]

I finally wrote this out in a semi-formal way. The documentation describing turning global gridded yield data into multipliers is a bit qualitative/descriptive, but the equations showing how to update the AgProdChange values using the multipliers and the original AgProdChange values is now typeset in a comprehensive way. And colorcoded.

[Miayuyao]

I'm trying to integrate GGCMI data into GCAM recently, but I haven't figured out how to generate the corresponding ag_pro change xml files. I would like to consult you if the useful code has been integrated into the latest version of gcamdata. Thank U~

[abigailsnyder]

Hello! The code for this is not currently in gcamdata. It relies on post-processing of specific files built by the GCAM data system (the L2052.AgProdChange* files). The impact multipliers for many of the AgMIP GGCMI FastTrack model output are available under the S1 data of this paper (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0237918#sec007). These have processed gridded data to correspond to the GCAM-basin level and GCAM commodities. The code to then turn those multipliers into updated AgProdChange xml files is not currently supported. We can provide a write up of how the multipliers get used to update AgProdChange values (attached), but again, these are not currently supported gcamdata features.

Impacts_GCAM.pdf

[Miayuyao]

Hello! The code for this is not currently in gcamdata. It relies on post-processing of specific files built by the GCAM data system (the L2052.AgProdChange* files). The impact multipliers for many of the AgMIP GGCMI FastTrack model output are available under the S1 data of this paper (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0237918#sec007). These have processed gridded data to correspond to the GCAM-basin level and GCAM commodities. The code to then turn those multipliers into updated AgProdChange xml files is not currently supported. We can provide a write up of how the multipliers get used to update AgProdChange values (attached), but again, these are not currently supported gcamdata features.

Impacts_GCAM.pdf

You are so nice! Thanks for sharing and your patience in explaining, I now try to follow the step-by-step approach on the pdf. But I have a problem that I don't have the shp file or other file format of aglu subregion division, I wonder if you can send me a copy? The GCAM version I'm trying to use is 4.4, so it's 283 sub-regions using AEZ partitions and 32 geopolitical zones.

[abigailsnyder]

I did some looking through our various repositories, and I don't think we currently offer support via shape files for the old AEZ version.