search

PNNL-TES / rcq10

1 stars 0 forks source link

Notes from talking with Peter, Xiaojuan, Dan #8

Open bpbond opened 5 years ago

bpbond commented 5 years ago

Dan is working on a Q10-related study w/ ELM.

Peter asks:

bpbond commented 5 years ago

FYI @jinshijian @thorntonpe @dmricciuto @xyk-ornl

Screen Shot 2019-05-01 at 9 32 29 AM Screen Shot 2019-05-01 at 9 32 40 AM
dmricciuto commented 5 years ago

Thanks for sharing these notes. Xiaojuan's GitHub is xyk-ornl.

bpbond commented 5 years ago

Could calculate, using existing model outputs, the apparent Q10, right? (Intrinsic is 1.5 here we know.) That is, compute a Q10 based on the autotrophic respiration response to temperature, combining substrate and temperature effects

Hi all, in particular Xiaojuan @xyk-ornl - @jinshijian and I are wondering, how much work would this be to do, from standard historical transient run? I.e., you guys extract soil surface CO2 flux (or whatever we need to calculate that), plus soil temperature at say 10 cm, for every timestep. We would then take these data and compute, for each day/night period, an "apparent Q10" (what would have been measured at the soil surface by some simulated researcher in the model world) by fitting R=ae^bT to the data.

We know you guys are really busy right now, so no pressure, just wondering. Thanks!

xyk-ornl commented 5 years ago

Hi @bpbond @jinshijian, it is pretty straightforward to output soil surface CO2 flux (soil respiration plus root respiration) and soil temperature for top 10cm at every time step. But doing that for the whole historical time period might be too much - taking long time to run and the output file would be huge (with half-hour time step). Maybe we can run only 1980-2010 and output these variables hourly - that would be more practical. Please let me know what you think. Thanks

bpbond commented 5 years ago

Oh gosh, agreed. I think 10 years (2001-2010) would be fine, outputted hourly. Would it be possible to add the photosynthesis->root C allocation flux too?

xyk-ornl commented 5 years ago

yes - we have that in the model and just need to add it to the hourly output list. Thanks

bpbond commented 5 years ago

So basically just a 10 (or even 5)-year uncoupled run outputting those variables...if not a huge amount of trouble (and absolutely at your convenience) that would be super. Thank you!

xyk-ornl commented 5 years ago

@bpbond @jinshijian

Just a heads up - After the CBGC paper and a few meeting and/or project related travels, I was back to working on this earlier this week. However I ran out of my NERSC allocation in the middle of the simulation - I was trying to run the transient simulation from 1850 to generate the 2000 initial condition. Since I will be out of office again starting next week, I've asked @dmricciuto to help us on this and hopefully we will have the outputs for you soon. I am sorry for the delay....

FYI, here are the variable list I put together

Root respiration

FROOT_MR LIVECROOT_MR CPOOL_FROOT_GR CPOOL_FROOT_STORAGE_GR TRANSFER_FROOT_GR CPOOL_LIVECROOT_GR CPOOL_LIVECROOT_STORAGE_GR TRANSFER_LIVECROOT_GR CPOOL_DEADCROOT_GR CPOOL_DEADCROOT_STORAGE_GR TRANSFER_DEADCROOT_GR

Soil respiration

HR

Root allocation from photosynthesis

CPOOL_TO_FROOTC CPOOL_TO_FROOTC_STORAGE CPOOL_TO_LIVECROOTC CPOOL_TO_LIVECROOTC_STORAGE CPOOL_TO_DEADCROOTC CPOOL_TO_DEADCROOTC_STORAGE

Soil temperature for top 10cm soil

TSOI_10CM

bpbond commented 5 years ago

That looks great--thanks @xyk-ornl @dmricciuto

dmricciuto commented 5 years ago

@bpbond @jinshijian I finished the runs on anvil but the outputs are huge - 1.6T. Do you have an account there? Or should I move to NERSC, or somewhere else?

bpbond commented 5 years ago

Uh wow, that's large. Yikes. @jinshijian we should talk to Sean to see if he has a multi-terabyte drive he could loan us.

I don't have an anvil account, but do on NERSC, so that would be easiest. Thanks Dan!

jinshijian commented 5 years ago

Wow, looks like you guys @xyk-ornl @dmricciuto have lots of progress on this！That is awesome, thank you so much. For some reason, my Gmail blocked the notification from Github, sorry for the delay response.

dmricciuto commented 5 years ago

@xyk-ornl @bpbond @jinshijian Sorry I haven't gotten back to you with the output! I finally transferred it over to NERSC, see here: Please see: /global/cscratch1/sd/ricciuto/e3sm_scratch/from_anvil - this includes 10 years of hourly outputs for the variables listed above. Let me know if you have any trouble accessing or if you need further explanation of the outputs.