search

JGCRI / gcam-core

GCAM -- The Global Change Analysis Model
http://jgcri.github.io/gcam-doc/
Other
282 stars 163 forks source link

I'd like to run the model each SSP scenarios. #345

Open smitsu7 opened 1 year ago

smitsu7 commented 1 year ago

Hello everyone. I'm a student of urban climatology lab. I have a question about how to run other SSP scenarios on GCAM. Actually, I try to modify all name of files which include ssp written on configuration file to other ssp. For example, I changed ../input/gcamdata/xml/socioeconomics_gSSP2.xml to ../input/gcamdata/xml/socioeconomics_gSSP1.xml. Is that correct way to conduct the model as other SSP? I think this way is just modifying Income elasticity each industry. I'm wondering whether this procedure is sufficient to explore other SSP scenarios.

If this way is wrong, could you tell me the correct one?

pkyle commented 11 months ago

I'm sorry for the delay in responding. The SSPs are set from a large number of XML files, many of which are not obviously associated with an SSP (e.g., not all scenario-specific XML files have the string "ssp" in the filename). The full suite of XML files per each SSP scenario is indicated in the batch configuration file batch_SSP_REF.xml, which uses configuration_ssp.xml as its base. If you save the latter as configuration.xml and run it, that will cause all 5 SSPs to run, so of course you may want to selectively delete some of the SSP scenarios from the batch file, and/or supplement with any policy files you wish to run.

smitsu7 commented 8 months ago

Thank you. I try it.

muxuxuxu1 commented 4 months ago

Thank you. I try it.

Hello, I would like to ask if you have successfully implemented the settings of different SSPs in GCAM.I would also like to try setting ssp4-4.5 in GCAM but the CO2 emissions I get are much lower than the data given in the CMIP. My setup is to set ssp4 and rcp4.5 in the configuration.xml and the rest is not changed, do you know what is wrong with this or do you know what to do to reproduce ssp4-4.5 correctly?

like this： batch_SSP_SPA4.xml

../input/policy/policy_target_4p5_spa4.xml
pkyle commented 4 months ago

To run SSP4-4.5, you'll want to open configuration_ssp.xml, change the BatchFileName from its current file to batch_SSP_SPA4.xml, and then in that file, at the bottom, delete the policy-target-runner rows for all but the 4.5 scenario. This will run the model iteratively in "target-finder" mode, so it should take hours, but in the end you'll get the 4.5 emissions trajectory. I'm not sure if that's what you've already done; just want to make sure that the instructions are clear. There's no guarantee that the results will be similar to the CMIP data, which is probably several years old at this point, and a number of model updates have taken place since then.

muxuxuxu1 commented 4 months ago

configuration_ssp.xml Thank you for your answer! I still have a small question, that is, how to set it in the configuration.xml to call the configuration_ssp.xml? Is it set in the run-gcam.bat, but when I set it like this, the model will report an error.

muxuxuxu1 commented 4 months ago

configuration_ssp.xml Thank you for your answer! I still have a small question, that is, how to set it in the configuration.xml to call the configuration_ssp.xml? Is it set in the run-gcam.bat, but when I set it like this, the model will report an error. What I did in the run-gcam.bat was to replace the last configuration.xml of the file with a configuration-ssp.xml, but then the runtime model would give an error. Is this because there are more files called in the configuration.xml than configuration-ssp? I would like to ask you how to call the configuration-ssp.xml correctly to achieve the purpose of setting different ssp-rcp in gcam? run-gcam.txt

muxuxuxu1 commented 3 months ago

I simulated GCAM SSP460 as you suggested, but the CO2 emission result (attachment 1) are different to both published data by The SSP4: A world of deepening inequality attchment1.docx attchment2.docx (attachment 2) , and online IMA dataset ( link:https://secure.iiasa.ac.at/web-apps/ene/SspDb/dsd?Action=htmlpage&page=about ). Namely, the carbon emission peak year and emission level at year 2099 are remarkably different.

I'm confused about such outcome. I also made some attempts for other SSP scenarios and such mismatch is still there. How can we explain such difference? Is there anyway that I can tune my GCAM model approximate to the published work. attchment1.docx attchment2.docx