Amazon Vegetation & Climate #26
Replies: 5 comments 4 replies
-
|
Potentially useful here are the regional climatology here for the |
Beta Was this translation helpful? Give feedback.
-
Can land parameters increase ET and precip over the Amazon?Investigated with LMWG_dev_#76 and 77.
Can better land initial conditions improve the situation?Investigated with #25
When did we start seeing this behavior?
The focus of #15 was on new ocean model tunings, but I think it is the first coupled model run that used CTSM5.3. This introduced new datasets, parameter file, and a new fire model. Changes are pretty ubiquitous, but reductions in tropical and boreal forest LAI seem consistent with aims for the PPE informed calibration of land-only runs introduced in CTSM5.3. Notably these decreases in Amazonian LAI are not associated with decreases in precipitation and only modest changes in air temperature Fire C losses do increase, but it seems unlikely fire alone is causing the drop in LAI?There's some indication that QBOT decreases over southern and eastern Amazon, were our our regional plots focus? |
Beta Was this translation helpful? Give feedback.
-
|
Some of this is review from the LMWG_dev discussion, but I'm confused by the
@adamrher, @dlawrenncar, @linniahawkins, @djk2120, @katiedagon, @aswann, @czarakas, @olyson, @slevis-lmwg |
Beta Was this translation helpful? Give feedback.
-
|
I agree that the fact that this behavior is not replicated in an F case is
odd. But, it likely isn’t coincidence that this is happening with first use
of ctsm5.3. Not sure whether it is unlikely or not that fire changes could
do this. A test would be to run 112 but reverting back to whatever CTSM was
being used in 104 or whichever is the most appropriate run. If the Amazon
problem goes away, then that is a big clue and we will have to try to
understand what looks like pretty thresholdy behavior.
…On Fri, Dec 6, 2024 at 5:08 PM will wieder ***@***.***> wrote:
Some of this is review from the LMWG_dev discussion, but I'm confused by
the
1. introduction of CTSM5.3 coinciding with this change in Amazon LAI
and our
2. inability to replicate the behavior in F-cases?
@adamrher <https://github.com/adamrher>, @dlawrenncar
<https://github.com/dlawrenncar>, @linniahawkins
<https://github.com/linniahawkins>, @djk2120 <https://github.com/djk2120>,
@katiedagon <https://github.com/katiedagon>, @aswann
<https://github.com/aswann>, @czarakas <https://github.com/czarakas>,
@olyson <https://github.com/olyson>, @slevis-lmwg
<https://github.com/slevis-lmwg>
—
Reply to this email directly, view it on GitHub
<#26 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AFABYVAPX533CB3226RD4WD2EI4A3AVCNFSM6AAAAABTFSPCPOVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCNBZGEZTKNQ>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
@wwieder @dlawrenncar I think it may be relatively simple for me to replicate the B112 case with the old fire scheme. Seems worth trying. I am more suspicious of the fire update than of the change to 5.3 datasets. But we can follow up with that right after if needed. |
Beta Was this translation helpful? Give feedback.
-
|
I started what I'm calling b112oldfire: NCAR/LMWG_dev#82 |
Beta Was this translation helpful? Give feedback.
-
|
Note, reverting to CLM5 fire does not save the amazon, although LAI in the E. Amazon does beneficially increase with the old fire scheme. See also this regional climatology |
Beta Was this translation helpful? Give feedback.
-
|
@slevis-lmwg Thanks. That's a lot of diffs between b112 and b111, but it does seem like fire is the most likely. I see that there are new params for P thresholds for fire in tropical forests (?). That looks like something that could be a factor and which could explain a somewhat thresholdy behavior between F cases and B cases. Perhaps there is just a smallish reduction in P in B cases and then that leads to some vegetation die off and that leads to less P and there is a feedback. If you look at F cases and B cases for P in Amazonia, it's quite a bit drier in B cases (before and after fire model changes), so I think it makes sense that we are seeing a response in B cases, but not F cases. This really points to the new fire model as being the culprit, but we will have confirmation shortly.
|
Beta Was this translation helpful? Give feedback.
-
|
One more note, the regional temperature and precip biases we're seeing in B-cases with CESM3 are worse that what we had in CESM1 and CESM2, land diagnostics here |
Beta Was this translation helpful? Give feedback.
-
In B cases we've noticed low leaf area index over the Amazon (compared to land only CRU-JRA runs, below).
We do have some suggestions from recent work by @czarakas and friends suggests that land model parameters can influence regional precipitation. How much can land model parameters help bring back rains (and clouds) over the Amazon?
Links above are compared to land only (I-cases), but we see similar differences between B and F-cases.
Summary of relevant LMWG diagnostics
Relevant Land Diagnostics
B-116 vs. I-5.3
B-116 vs. B-112
B-112 vs. B-111
B-116 vs. F-0008, the F-case is a CLM6-SP case
Lots of LMWG diagnostics posted from B cases here (thanks @olyson)
Notes on the dry amazon bias
Notes on low LAI biases:
We identified this came in with #15, which also included updates to CLM5.3 which included a new
Fire Model: investigated with #82 in LMWG_dev
Land parameterization: investigated with #85 in LMWG_dev
Surface Data, not assumed to be responsible for this issue
Additional discussion can also be found on the LMWG_dev page.
Beta Was this translation helpful? Give feedback.
All reactions