General Comments

This contribution assesses the benefits of negative emissions / CO2 removal technologies deployment for future climate states using an ensemble of intermediate complexity earth system model results. The authors use effective transient climate response to cumulative CO₂ emissions (eTCRE) and zero emissions commitment (ZEC) as metrics to quantify these impacts. The authors find that thermal dependance and airborne fraction of CO2 contribute almost equally to the uncertainty in eTCRE, which is in contrast with recent analysis of the CMIP6 ensemble.  Additionally, the authors find that negative emissions deployment can help avoid continued warming after net-zero emissions are reached.  The manuscript is clear and well-written and the analysis appears free of errors.  However I have several recommendations aimed at increasing the impact and clarity of this work, which are detailed below.   

Specific Comments

The authors use the RCP 4.5 medium-level mitigation scenario as a benchmark to assess the future climate response to negative CO2 emissions. However much of the discussion of prospective large-scale negative emissions deployment in the recent literature focuses on their use towards limiting end-of-century warming to well-below 2 C, more consistent with RCP 2.6 or RCP 1.9. Although even the “medium” mitigation scenario may seem optimistic relative to the present real-world trajectory, using one or both of these forcing scenarios representing even deeper levels of mitigation could increase the impact of this work. I recommend the authors run similar analysis on one or both of these deeper mitigation scenarios, even as a sensitivity case. This could allow the modeling community, policymakers, and other stakeholders insight into what a “best case” scenario might look like in terms of transient climate response and committed warming. 

In the final paragraph of the conclusions section the authors refer to the need for negative emissions technologies that have naturally long storage times.  In the main body of the manuscript it would be helpful to describe exactly what types of carbon removal technologies are represented in the models used to develop the ensemble.  Long-lived and permanent storage such as direct air capture or enhanced rock weathering referred to in the conclusions?  Or biospheric such as afforestation? Or is the representation of negative emissions agnostic as to the source in the models?   Any biospheric contribution to the negative emissions and potential feedbacks or the limitations in representing them should be identified in the discussion around land carbon. 

Technical Corrections

The authors should be more precise in differentiating point source carbon capture from carbon capture from the atmosphere for negative emissions. Throughout the manuscript the more generic “carbon capture” or “carbon capture and storage” is used.  While readers might infer from context that this is referring to negative emissions, this term should be clearly defined at every use to avoid the possibility of misinterpretation. 

Review of: Assessment of negative and positive CO2 emissions on global warming metrics using

large ensemble Earth system model simulations

Vakilifard et al.

Overall Assessment:

The study uses an intermediate complexity Earth system model to assess eTCRE, ZEC and the effect of net negative emissions on these metrics. In addition the study uses the theoretical framework of Williams et al., 2016 to decompose TCRE into components. Although the perturbed parameter model set-up is a promising method, the experiment design is baffling and thus the wider meaning of the results is very unclear. Additionally the paper focuses on far too many metrics leading to a manuscript that is exceptionally long but without discernible purpose or meaningful conclusions. I recommend that the paper undergo major revisions.  

General Comments:

(1) The authors need to explain where the 82 (86?) variants of GENIE come from. There is a reference to Foley et al., 2016) which presumably first derived the model set-up but a paper of this length should be self-contained and explain to the reader which parameters are being perturbed and how the variants represent uncertainty in the climate system.

(2) The experiment design does not make much sense. The design is an example of a half-idealized experiment, as the non-CO2 forcings are frozen at 2020 values but there is no explanation as to why this is done. Why not follow the real RCP 4.5 pathway? There are times when half-idealized experiments do make sense but their purpose needs to be clear which is absent here. Also why use RCP 4.5 instead of SSP 4.5? What is gained by starting the experiments in 850 CE instead of 1850 CE?  The relevance model simulations would be easier to interpret if either real scenarios or idealized experiments were conducted. 

(3) Following from the strange experiment design the computation of eTCRE if not comparable to other models and the computation of ZEC is incorrect. eTCRE by its nature varies by the pathway of non-CO2 forcing agents. Here the effect of these agents is frozen in 2020, meaning that eTCRE cannot be compared to eTCRE computed by other models for RCP 4.5. 

ZEC is defined as the Zero Emissions Commitment and can be defined as CO2-only ZEC, or ZEC for other forcing agents independently or in combination (e.g. Matthews & Zickfeld 2012). Here none of these protocols is followed and you attempt to compute ZEC for zero CO2 emissions while holding the non-CO2 forcings constant. Thus the metric you compute is not ZEC, nor can it be compared to ZEC derived from other model simulations.

(4) While the decomposition of eTCRE into components is fine it is unclear what this analysis adds to our understanding of eTCRE. The paper essentially does the decomposition explains the results but does not assess how the 28 perturbed parameters are controlling the evolution of these components.  

Specific Comments:

Line 2: K EgC^-1 is the preferred unit. Having Celsius and carbon together is confusing. 

Line 180: Does this mean that 4 of the model variants crashed?

Figure 2: Grey lines for each variant with a mean value would be easier to look at.

Figure 3: The release of carbon from sediments seems really high.

Line 224: Does Genie have any other heat sinks? Estimate is wrong word, makes it seem like energy is not being conserved by the model.

Figure 10: Needs work. Different line styles for the percentiles would improve legibility.

References:

Matthews, H. Damon, and Kirsten Zickfeld. "Climate response to zeroed emissions of greenhouse gases and aerosols." Nature Climate Change 2.5 (2012): 338-341.