On linking an Earth system model to the equilibrium carbon representation of an economically optimizing land use model

Bond-Lamberty, B.; Calvin, K.; Jones, A. D.; Mao, J.; Patel, P.; Shi, X. Y.; Thomson, A.; Thornton, P.; Zhou, Y.

doi:https://doi.org/10.5194/gmd-7-2545-2014

Articles | Volume 7, issue 6

Geosci. Model Dev., 7, 2545–2555, 2014

https://doi.org/10.5194/gmd-7-2545-2014

© Author(s) 2014. This work is distributed under
the Creative Commons Attribution 3.0 License.

Geosci. Model Dev., 7, 2545–2555, 2014

https://doi.org/10.5194/gmd-7-2545-2014

© Author(s) 2014. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 7, issue 6

Development and technical paper 06 Nov 2014

Development and technical paper | 06 Nov 2014

On linking an Earth system model to the equilibrium carbon representation of an economically optimizing land use model

B. Bond-Lamberty et al.

Download

Final revised paper (published on 06 Nov 2014)
Preprint (discussion started on 27 Feb 2014)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC C258: 'Difficult to understand', Anonymous Referee #1, 04 Apr 2014
- AC C829: 'Response to Referee #1 comments', Ben Bond-Lamberty, 07 Jun 2014
RC C491: 'Design and modelling choices unclear', Anonymous Referee #2, 06 May 2014
- AC C838: 'Response to Referee #2', Ben Bond-Lamberty, 07 Jun 2014

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Ben Bond-Lamberty on behalf of the Authors (07 Jun 2014) Author's response Manuscript

ED: Referee Nomination & Report Request started (11 Jun 2014) by Patrick Jöckel

RR by Anonymous Referee #2 (29 Jun 2014)

Suggestions for revision or reasons for rejection

Second review of manuscript by Bond-Lamberty et al: „On linking an earth system model to the equilibrium carbon representation of an economically optimizing land use model“

The authors have presented a manuscript which has significantly improved since the initial submission. While I’m generally happy with the direction this is going, I hereby suggest a few final additions. These additions should not require much additional work and time, but would help better communicate the strengths and weaknesses of the presented approach.

First of all, I think the title is an excellent choice and now better prepares the reader for what is coming.

Further comments:
Page 14, formulas (1) and (2): C_{A} and C_{B} would need an additional index to indicate the time step
Page 15, line 336: I don’t see where the 20-92%comes from. Looking at the R squares Table 2, I can see a range of 6 to 92%. Please make sure to correct, and also not to round only the lower end in order to make things look nicer than they potentially are.
Page 16, line 347: Please include by how much (in relative terms) the regression slopes changes. Are we speaking about a 1, 5, or 50% change in the slope?
Page 16, line 361: From figure 5, I cannot see where the “<10%” is coming from. For perturbations of 25% of the cells or less, I would say there is a maximum error of about 20%. Please correct.
Page 17, lines 381-387: This discussion of caveats should explicitly include the limitation that this approach has been tested for a scenario with modest climate change (RCP4.5). It’s validity for high climate change futures is still to be tested.
Page 34, Figure 2 – caption: As far as my eyes are not tricking me, the acronyms are still not consistent, even not between caption and figure. Please go over this once again.
Page 35, Figure 3: Please also show the result for the final design decision and methodology.
Page 36, Figure 4: Please color the points that were excluded due to the outlier test.
Page 37, Figure 5: I do not understand what the isoline labels are trying to indicate, as there appear to be isolines every 20% and the caption indicates that they show errors for every 25%.
On the argumentation for the choice of RCP4.5 vs 8.5: The only valid argument here is probably that RCP8.5 would have involved more work. Both RCP4.5 and RCP8.5 are part of the first tier of CMIP5 experiments and have thus been run by all climate modelling groups. Moreover, for RCP8.5 a large perturbed initial condition ensemble is available from CESM, which arguably would be able to provide even more interesting information into the importance and influence of internal variability of a complex AOGCM on the uncertainties of the relationships suggested here.

Small editorial comments:
Page 4, line 71: It is better to cite more recent papers which include the model description.
Meinshausen, M., et al. (2011). "Emulating coupled atmosphere-ocean and carbon cycle models with a simpler model, MAGICC6 – Part 1: Model description and calibration." Atmos. Chem. Phys. 11(4): 1417-1456.
Meinshausen, M., et al. (2011). "Emulating atmosphere-ocean and carbon cycle models with a simpler model, MAGICC6 – Part 2: Applications." Atmos. Chem. Phys. 11(4): 1457-1471.
Page 7, line 140: No “-“ in land use.
Page 9, line 200: comma after “two”

Hide

RR by Anonymous Referee #1 (08 Aug 2014)

Suggestions for revision or reasons for rejection

The manuscript is in in better shape than before but it still needs some revisions. Abstract still needs improvements. It doesn’t deliver the primary message clearly and in fact contact few incorrect statements. The overall “loose” coupling of CLM and GCAM appears so “non-scientific”. I am not sure if it is the way it is described or if it actually is, probably the former.

Abstract. Only on reading the manuscript the second time it was obvious to me what authors meant by “… which can incorporate land-use change as static inputs but do not simulate the policy or economic forces …”. This sentence still doesn’t make sense. How can land-use be “changing” and still “static”. What authors mean here is likely “prescribed”. That land-use change is prescribed makes sense. I still can’t appreciate “but do not simulate the policy or economic forces”. This needs to be better worded.

Abstract. “ … focusing on how CESM climate effects on the carbon cycle can be shared with GCAM”. No, this can’t be your focus since this is an offline study.

Abstract. “ … quantify how carbon flux changes driven by climate (e.g. CO2 fertilization) …”. CO2 fertilization is NOT a climate effect.

Abstract. “ … to manipulate GCAM’s assumptions of equilibrium ecosystem steady state carbon”. Please consider using another word for “manipulate”.

Line 55. “ … different policy choices vis-à-vis LUC and carbon may result in greatly different configurations of the future carbon cycle”. Still a vague sentence.

Lines 64-66. “Such models may incorporate LUC as static inputs, but do not simulate policy options or economic forces, a significant limitation given how strongly humans can perturb the earth system”. See comments above regarding “static” and “simulate policy options or economic forces”.

Line 99. “CLM simulates the cycling and land-atmosphere exchange of energy, water, chemical elements, and trace gases”. CLM simulates fluxes of CO2 and N2O, as far as I know. “land-atmosphere exchange of chemical elements” sounds confusing. Please be explicit here.

Line 122. Market-clearing prices is a term unknown to ESM community.

Line 126. What is a logit mathematical model?

Line 130. “GCAM’s carbon model”. Wouldn’t a more appropriate term be “GCAM’s terrestrial carbon cycle model”.

Line 137 and 140. “When a particular land-use contracts” and “When a land-use expands”. Wouldn’t it be appropriate to tell your reader what kind of land-uses GCAM deals with to provide some context.

Line 168-169. “have a suitable proxy by which to change GCAM’s steady-state carbon assumptions”. You are not changing GCAM’s steady state carbon assumptions (it still assumes that pools are steady state) but rather the magnitude of the pools which GCAM assumes are in steady state. Please also address this subtlety elsewhere in the manuscript.

Line 171. “ … if the carbon stock of a CLM forest changes from one time step …”. Please reword to “ … if the land carbon pool size of a grid cell with forested fraction simulated by CLM changes …”

Line 205. “In simulations S2-S4, we used the same looped climate, …”. Reword to “In transient 1850-2010 S2-S4 simulations, we used the 1901-1920 climate repeatedly …”

Line 206, 207. Replace “factors” by “forcings” and elsewhere too.

Line 237. “… even a perfect proxy variable will be subject to climate and LUC …”. You mean subject to “climate change and variability”. “subject to climate" doesn’t mean anything.

Line 258. “To distinguish LUC from climate signals …”. Doesn’t this actually mean, “To distinguish the effect of LUC, on primary CO2 fluxes and land carbon pools, from that of climate …”

Line 267. “The scalars were then …” What scalars are being referred to here?

Line 300-302. “ … because a few CLM grid cells, located in GCAM’s “Middle East” region, were subject to LUC at the end of CLM’s spinup phase”. This is confusing for me as an ESM person. Typically a spin up is performed when all forcings are fixed. For example, a typical ESM control run spinup is done with land cover fixed at 1850 conditions and concentration of all GHGs at their 1850 values. In the context of this paper the spinup S1 is done with 1901-1920 climate used repeatedly. Table 1 doesn’t say anything about land cover so I presume it would be specified and not changing but the above sentence implies that land cover is changing. If land cover is changing in a spin up simulation then how come it is a spin up.

Equations (1) and (2). Shouldn’t equation (1) be more appropriately written as
CA_i = CA_0 x (NPP_i/NPP_0)

In regards to equation (2), isn’t a more appropriate quantity to describe change in the belowground carbon pool (NPP – HR)_i/(NPP – HR)_0, as opposed to the average of the NPP_i/NPP_0 and HR_i/HR_0 ratios. After all the rate of change of soil carbon is equal to NPP minus HR, assuming NPP equals litterfall rate. In a transient simulation NPP increases faster than HR, so the difference between the two will reflect how soil C pool is changing.

Also, I presume, equations (1) and (2) are applied every 5 years.

Line 336. “The initial change in NPP was able to explain 20-92% of the variance …”. You mean, “The change in NPP over the 2005-2009 period was able …”

Lines 348-353. These sentences are redundant and provide no useful information.
Section 3.3. Wouldn’t a more appropriate title be “Distinguishing the effects from LUC from climate”

Line 360. “ … that as long as fewer than ~25% of the simulation cells were perturbed, the error (between the known climate signal and that inferred by the outlier test) ..” What exactly is this error? What are its units? What are the units of the “climate signal”? I am sorry but I am unable to understand what error is being talked about here.

Line 375. “Climate changes” doesn’t seem right. It’s either “Climatic changes” or “Climate change”.

Line 377,378. “In addition, it guarantees that if climate changes affect the carbon cycle, GCAM’s equilibrium assumptions will change correspondingly for the same vegetation type and spatial location, feeding back into economic decisions about industrial and LUC CO2 emissions (e.g., Le Page et al., 2013), emissions that propagate back to CLM (Di Vittorio et al., 2014)”. Are you sure, this is incorrect. I don’t think LUC CO2 emissions from GCAM propagate back to CLM. It’s the land use information, which is then translated to land cover information that is seen by CLM.

Lines 384, 385. “In addition, the outlier-exclusion step will break down under extreme LUC scenarios, scenarios that can be a useful tool”. I am unable to understand the last part of this sentence – “scenarios that can be a useful tool” – how does this relates to the first part of the sentence.

Line 400. “This work is only one step to a full coupling of an ESM and IAM; the second is described by Di Vittorio et al. (Di Vittorio et al., 2014).” Either don’t mention the second step or if you do want to mention it then please briefly explain what is it.

Figure 5. I am unable to understand Figure 5. Like I mentioned earlier, what exactly is error. What are its units? The caption says “error between the inferred climate change signal and the known (artificial) signal”. But what is “climate change signal”? Caption says x-axis is ‘LUC intensity” and the figure itself says “LUC effect”.

Hide

RR by Anonymous Referee #3 (30 Aug 2014)

Suggestions for revision or reasons for rejection

This manuscript describes a loose coupling between an ESM and an IAM as a step towards a fully coupled implementation of an integrated ESM/IAM. Here, they focus on one particular challenge in this coupling, namely that the GCAM IAM model typically assumes constant carbon stocks but the ESM predicts time-varying carbon pools. This inconsistency can be rectified if the equilibrium carbon stocks utilized in GCAM could be updated based on information from the ESM. The challenge is that one needs to separate the changes in carbon stocks due to climate change and CO2/N fertilization from the changes due to LUC. The authors show that the equilibrium carbon stocks in GCAM can reliably be adjusted by scaling by changes in carbon fluxes (NPP and respiration) from the ESM, as long as rapid shifts in fluxes, which are typically due to LUC are eliminated.

Overall, I found the paper to be concise and well-written. The group that is trying to couple CESM to GCAM should be commended for their systematic approach to this challenging problem. This paper represents an important step both from a conceptual point of view in terms of how to complete a loose coupling and from an understanding point of view with respect to how the carbon cycle operates in CLM and in GCAM.

I don’t have many criticisms of the paper. I find it suitable for publication with minor revisions. Some suggested changes are included below.

1. p. 10, line 227: In these constant LUC experiments, how is wood harvest treated. In fact, it wasn’t clear how the impact of wood harvest was treated/considered throughout the paper. Wood harvest will change the stocks and may affect the climate change response, but may not result in a big shift in fluxes so could pass through the outlier exclusion test (?). Maybe this is fine, but some discussion is probably warranted.
2. Line 282: Does this revised text better represent what you are trying to say here: “this inferred climate change induced carbon stock signal was then compared to the original known climate change induced carbon stock signal.” If so, maybe change the wording to clarify.
3. Equations 1 and 2: I think it would be clearer to include a subscript for CA and CB. E.g. CAi = CA0(NPPi/NPP0). I am assuming that CA0, is the initial equilibrium carbon stock based on Houghton and that at each timestep i, this carbon stock is adjusted to CAi.
4. Along the same vein, what exactly do you mean by timestep here? I think that you mean the IAM timestep of 5 years. Does NPPi represent the average NPP over the prior 5 years? What exactly does NPP0 represent?
5. Line 401: Should describe in a sentence or two the second step that is described in Di Vittorio et al.

Hide

ED: Publish subject to minor revisions (Editor review) (11 Sep 2014) by Patrick Jöckel

AR by Svenja Lange on behalf of the Authors (25 Sep 2014) Author's response

ED: Publish subject to technical corrections (29 Sep 2014) by Patrick Jöckel

AR by Ben Bond-Lamberty on behalf of the Authors (29 Sep 2014) Author's response Manuscript