General comment:

This study analyzed the PDRMIP multi-model ensemble to show that changes to the surface energy budget are distinctively different in the case of the BC-induced climate change from those enforced by other forcing agents such as CO2, scattering aerosol, and solar insolation. Specifically, the authors show that the climate response of the surface air temperature is governed by changes to multiple components of the surface energy budget in the BC case, contrary to other scenarios where surface radiative flux changes is the major factor that mostly determines the temperature response. I think this study is a nice follow up for recent studies that similarly assessed the surface energy budget change for some of the forcing agents to identify their different characteristics of climate responses. I would recommend the paper be considered for publication in ACP contingent upon if the authors appropriately address my specific concerns described below.

Specific comment:

Line 34-39: The numbers listed here should be shown with appropriate units (dimensions) since these numbers are for a unit BC forcing. For example, the energy flux changes (radiative, sensible and latent) should be dimensionless (rather than having the unit of Wm^-2) given that they are normalized by the TOA radiative flux change.

Line 38-39: I’m a bit surprised that the sensible heat flux change (2.53Wm^-2) is larger than the latent heat flux change (1.30Wm^-2) simply because of my naïve understanding that the latent heat typically dominates the turbulent heat transfer from the surface to atmosphere. Can the authors explain why the opposite (i.e. sensible heat is larger than latent heat) occurs in the BC-forced scenario? The enhanced stability just explains the sum of latent and sensible heat changes but no explanation for their partitioning.

Section 2.2: I’m wondering if the energy balance described by equations (2) and (3) indeed applies when analysis is restricted to land grids only. Don’t these equations need additional terms for energy exchange between land and ocean? Did the authors confirm that the balance relationship of (3) is indeed true in the model data analyzed? This can possibly be addressed by adding a bar for the residual in Fig. 4.

Minor point:

Line 101: relative -> relatively

Line 205: “wind speed (U) is likely the only main factor driving the change of latent heat flux”: How about the specific humidity (qa)? The enhanced stability in the BC case can also change qa.

Line 256: I don’t understand what ‘bottom-up’ means here. To my understanding, the energy balance constraint discussed throughout this paper is all ‘top-down’ regardless whether it is for top-of-atmosphere or surface. Why is it called ‘bottom-up’ when the surface energy response to solar insolation change is discussed? Please clarify.

This paper explores the climate response at the surface to BC aerosol forcing in a set of sensitivity tests with a variety of global climate models. The paper draws a contrast between the surface response to BC forcing and the surface response to a variety of other climate forcing agents, including greenhouse gases, solar variations, and mostly scattering sulfate aerosols. The paper argues that the response at the surface to the attenuation of downward solar radiation attributable to BC aerosols elicits a significantly stronger compensating response in the surface turbulent fluxes compared to the other forcing agents. This is attributed to the increased static stability and closely related reduction in surface winds that the authors argue is more significant in response to BC aerosol forcing compared to the other forcing agents. The paper is useful contribution and the approach is fairly robust, apart from the usual limitations of global climate models. The paper is suitable for publication following some minor revisions.

Major comments:

The individual figure elements in figures 2, 3, 5, 6 and 7 are very small. Presumably the purpose of showing a global map would be to reveal the regional differences evident in the map, only some of which are discussed in the text. Nevertheless, such variations are difficult to discern in maps as small as included here. If the intent is for the reader to simply zoom in on a computer screen, then I suppose it is sufficient. This is perhaps an editorial decision regarding whether the figures must be accessible to the reader in print form, or whether it is sufficient to presume the reader will zoom in electronically. Nevertheless, much of the argument rests on what are essentially global energy budget arguments with only limited discussion of the regional differences apparent across what amounts to more than 60 global maps. The authors could consider conveying more of the quantitative results in simpler bar figures, such as figure 4 and including global maps only for key results where the regional variations are central to the argument.

The results shown in figures 2 and 3 describe the basic physics underlying the main point of the paper, but only for energy budget considerations over the land areas. Of course, from a global perspective, the total energy balance is strongly determined by the ocean response. The authors acknowledge that they are only focusing on land, but this choice is not justified in the discussion. Perhaps the notion is that the temperature response addressed in the following figures is more important to people over land because that is where the people are. If so, then the authors should state it, since the choice to ignore the ocean when considering surface energy balance responses to forcing agents is not intuitive.

The differences between the various forcing experiments shown in figure 5 look very dramatic. In particular, the BC experiment stands out strongly in comparison with the other forcing agents, which of course is part of the point of the paper. However, how much of that is simply because the surface forcing per unit of TOA forcing is greater for BC? Put another way, if the solar forcing were reduced by enough that the reduction in solar insolation at the surface were comparable to the attenuation of 10x BC, would there be important responses in the turbulent fluxes that are not evident in figure 5? Typically, forcing from reductions in solar forcing, or forcing from purely scattering aerosols are of a similar magnitude at the surface as they are at the top-of-atmosphere. However, the surface forcing of a perturbation in BC at the surface can be several times that of the TOA forcing (see e.g. Magi et al. 2008 where the surface forcing efficiency of smoke can be approximately 10x the forcing efficiency at top-of-atmosphere).

In the abstract and discussion, the authors refer to a “top-down” influence of BC aerosols. I think I get what the authors are trying to imply by “top-down”, but I find the explanation to be rather vague. For example, static stability is typically quantified by a potential temperature gradient in the lower troposphere. That is above the surface, I suppose, but not the top of the atmosphere, or even top of the troposphere. I think that if the authors are going to include this notion, especially in the abstract, they need to be quite a bit more specific about what exactly is defining the volume or geometry of the space they consider having a “top” and “bottom” and why a term that implies a downward action is appropriate. In my opinion it does not make the physical argument any clearer.

Finally, while I find that the robust results from an ensemble of simulations comparing responses to a range of individual forcing factors to be a valuable contribution to the literature, I find the notion that “a clear and detailed mechanism of surface radiative response to BC is still lacking” (line 76) to perhaps be overstating the level of ignorance. While not purely a study of BC aerosol forcing, I would refer the authors to Liepert et al. (2004) for an early introduction to the notion that aerosol forcing at the surface can yield turbulent flux responses that are important on global scales.

Minor comments:

All of the quantitative forcing values noted in the abstract have positive signs, even while most of them are meant to quantify reductions in the forcing. Typically, reductions in a forcing value should be given a negative sign. Alternatively, and perhaps more rigorously, since many of the forcing quantities are directional (upward or downward) the authors could define changes that contribute an increase in the net downward forcing as positive in sign and decreases in net downward forcing as negative.

In line 181 I think “and all grid were given” should read “and all grid cells were given”.

Reference:

Magi, B.I., Fu, Q., Redemann, J. and Schmid, B., 2008. Using aircraft measurements to estimate the magnitude and uncertainty of the shortwave direct radiative forcing of southern African biomass burning aerosol. Journal of Geophysical Research: Atmospheres, 113(D5).

Liepert, B.G., Feichter, J., Lohmann, U. and Roeckner, E., 2004. Can aerosols spin down the water cycle in a warmer and moister world?. Geophysical Research Letters, 31(6).