The authors used a suit of models to explore the potential impact of solar engineering on permafrost carbon dynamics. They mainly found that the permafrost carbon feedback will be greatly inhibited in the scenario of solar engineering, and the permafrost ecosystems remains a carbon sink and the solar engineering will delay the transition of ecosystem carbon cycle from sink to source.  

First, the historical simulations have a large deviation from the observed permafrost in terms of area and active layer thickness. The model evaluation needs to be performed in a great detail. Such model deficiency could definitely introduce significant biases into our understanding of the solar engineering impact on the permafrost dynamics. Whether these biases would undermine the main results and conclusions are yet to be determined. The authors should explore the model biases and their potential sources, and discuss the potential impact of these biases on the results. Such model biases would also be present in the simulated terrestrial carbon components such as NPP and ecosystem respiration.

Second, I strongly disagree with the use of permafrost carbon throughout the manuscript. Almost all of the models did not really consider the permafrost carbon processes (freezing impact on decomposition, cryoturbation etc). Moreover, the models could even fail to capture the magnitude of permafrost carbon storage. The use of observation soil carbon data in this study did not consider the yedoma carbon pool, and the results presented here would therefore definitely mislead the readers. All of the expression related to permafrost carbon should be removed, since this is nothing to do with the real permafrost carbon dynamic processes. 

The authors studied how the permafrost extent and terrestrial carbon fluxes and pools are in response to two solar geoengineering scenarios, ssp245 and 585 scenarios based on CMIP6 simulations. The topic is timely and fits the scope of the journal Earth System Dynamics. The paper is well structured and the results are well illustrated in the figures. However, the authors need to put more effort to explain how the difference in surface climate between solar geoengineering scenarios and ssp245 affects soil surface temperature, NPP, and ER. Vegetation response to climate forcing plays a major role here. It is strange that G6solar and G6sulfur have lower radiation or summer temperature than ssp245, but they still have higher NPP and ER. Also, snow duration and surface litter are important insulators for soil freezing and thawing. I suggest the authors should mention these aspects in their discussion.  Finally, the authors should polish the language further and avoid some grammar mistakes.

Specific comments:

Title: to be more specific, like “two solar geoengineering scenarios”

Abstract:

Line 18, reduce -> reduces,

Line 18, “including” makes readers confused

Line 20-23: I suggest the authors describe more clearly what four scenarios are.  Are they two solar geoengineering scenarios based on the settings of ssp245 and ssp585 respectively?

Line 25 to 30: to report results in a quantitative way.

Introduction:

Line 35: “driven by Arctic amplification”, this sentence has a logical mistake. More rapid warming in the north than in the south actually means Arctic amplification.

Line 36: renders -> render

Data and methods:

Line 118: Is the G6sulfur run based on the setting of ssp585?

Line 122: You need to give some arguments on why the baseline period uses 20 years but the future period uses 10 years.

Deriving permafrost extent and ALT

Line 163 and 165: How did you linearly interpolate the soil temperature from the surface? The permafrost model should be able to simulate multi-layer soil temperature.

Results:

I suggest the authors add a table or time series for permafrost areal extent, which is complementary to figure 2.

In Figure 3 and Figure 4, I think the authors need to explain why G6solar and G6sulfur have lower radiation and summer temperature but still higher NPP and RH, compared to ssp245.

Figure 6. In the caption, “G6solar, G6sulfur, and ssp585” need to be reversed in their order.

You should present a figure of ensemble standard deviations from each scenario in the appendix.