• Alexander, M. A., , J. D. Scott, , and C. Deser, 2000: Processes that influence sea surface temperature and ocean mixed layer depth variability in a coupled model. J. Geophys. Res., 105, 16 82316 842.

    • Search Google Scholar
    • Export Citation
  • Compo, G. P., , and P. D. Sardeshmukh, 2010: Removing ENSO-related variations from the climate record. J. Climate, 23, 19571978.

  • Dommenget, D., 2009: The ocean’s role in continental climate variability and change. J. Climate, 22, 49394952.

  • Dommenget, D., , and M. Latif, 2002: Analysis of observed and simulated SST spectra in the midlatitudes. Climate Dyn., 19, 277288.

  • Dommenget, D., , and M. Latif, 2008: Generation of hyper climate modes. Geophys. Res. Lett., 35, L02706, doi:10.1029/2007GL031087.

  • Forster, P. M., , M. Blackburn, , R. Glover, , and K. P. Shine, 2000: An examination of climate sensitivity for idealised climate change experiments in an intermediate general circulation model. Climate Dyn., 16, 833849.

    • Search Google Scholar
    • Export Citation
  • Lambert, F. H., , M. J. Webb, , and M. M. Joshi, 2011: The relationship between land–ocean surface temperature contrast and radiative forcing. J. Climate, 24, 32393256.

    • Search Google Scholar
    • Export Citation
  • Roeckner, E., and Coauthors, 2003: The atmospheric general circulation model ECHAM5. Part I: Model description. MPI Rep. 349, 127 pp.

  • View in gallery

    SST response to Tland + 1 K forcing in different experiments.

  • View in gallery

    Tocean response to Tland + 1 K forcing in different experiments.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 4 4 2
PDF Downloads 2 2 0

Comments on “The Relationship between Land–Ocean Surface Temperature Contrast and Radiative Forcing”

View More View Less
  • 1 School of Mathematical Sciences, Monash University, Clayton, Victoria, Australia
© Get Permissions
Full access

Abstract

In a recent article, Dommenget discussed the role of sea surface temperature variability for continental climate variability and change. Lambert et al. comment on Dommenget’s article in their article several times, arguing that the sensitivity experiment in Dommenget, in which the SST response to surface land temperature changes are discussed, is inconsistent with their and other previously published studies. In this comment, the results of Dommenget’s sensitivity experiments are discussed in more detail and the experiments are extended for longer response times. It is shown that the discussion of how the oceans’ response to land forcing is time-scale dependent, with a very weak response to land forcing on interannual time scales, as discussed in Dommenget, and that it has about a twice as strong of a near-equilibrium response to land forcing on time scales longer than 100 yr. The asymmetric land–sea interaction, with the ocean forcing the land much more strongly than the land forces the oceans, as discussed in Dommenget, is confirmed by this study.

Corresponding author address: Dietmar Dommenget, School of Mathematical Sciences, Monash University, Clayton VIC 3800, Australia. E-mail: dietmar.dommenget@monash.au

The original article that was the subject of this comment/reply can be found at http://journals.ametsoc.org/doi/abs/10.1175/2011JCLI3893.1.

Abstract

In a recent article, Dommenget discussed the role of sea surface temperature variability for continental climate variability and change. Lambert et al. comment on Dommenget’s article in their article several times, arguing that the sensitivity experiment in Dommenget, in which the SST response to surface land temperature changes are discussed, is inconsistent with their and other previously published studies. In this comment, the results of Dommenget’s sensitivity experiments are discussed in more detail and the experiments are extended for longer response times. It is shown that the discussion of how the oceans’ response to land forcing is time-scale dependent, with a very weak response to land forcing on interannual time scales, as discussed in Dommenget, and that it has about a twice as strong of a near-equilibrium response to land forcing on time scales longer than 100 yr. The asymmetric land–sea interaction, with the ocean forcing the land much more strongly than the land forces the oceans, as discussed in Dommenget, is confirmed by this study.

Corresponding author address: Dietmar Dommenget, School of Mathematical Sciences, Monash University, Clayton VIC 3800, Australia. E-mail: dietmar.dommenget@monash.au

The original article that was the subject of this comment/reply can be found at http://journals.ametsoc.org/doi/abs/10.1175/2011JCLI3893.1.

1. Introduction

In a recent article, Dommenget (2009, hereafter D09) discussed the role of sea surface temperature (SST) variability for continental climate variability and change. The focus of this article is on the role of the natural SST variability in forcing interannual natural climate variability in continental surface temperatures. But some discussion in D09 is also focused on climate change scenarios and on the role of the ocean in the continental response to doubling of CO2.

Lambert et al. (2011, hereafter LWJ11) comment on D09 in their article several times, arguing that the sensitivity experiment in D09, in which the SST response to surface land temperature changes are discussed, is inconsistent with their and other previously published studies. In particular, LWJ11 point out that the simulation is only 20 yr long and is therefore unlikely to be in equilibrium. They further point out that the weak SST response is inconsistent with Forsters et al. (2000). LWJ11 implicitly argue (H. Lambert 2011, personal communication) that the experiment should result in the same land–sea warming ratio of 1.7 as in the OZ−2×CO2 experiment of D09.

Indeed, it seems reasonable to assume that the 20-yr SST response simulation in D09 is not in equilibrium. In this comment the results of the D09 sensitivity experiments are discussed in more detail, and it is further shown how they relate to long-time near-equilibrium responses. The experiment with land forcing is extended for a longer response time and compared with a twin experiment using a slab ocean model. Further, the simple toy model of D09 is used to discuss the relative roles of ocean and land in climate change, natural variability, and different sensitivity experiments to better understand the results of D09 and how they relate to other studies and that of LWJ11.

2. Model simulations

As in D09, all the simulations are based on the atmospheric GCM ECHAM5 (Roeckner et al. 2003) with a horizontal resolution of T31 (3.75° × 3.75°) and 19 vertical levels. In the OZ–TLAND experiment, the atmosphere is coupled to the simple 1D ocean mixed-layer model OZ (Dommenget and Latif 2008) as in D09. As described in D09, the ocean model OZ has 19 vertical layers that are connected through vertical diffusion only. It is further important to note that the OZ model’s lowest level is coupled, by Newtonian damping, to a fixed deep-ocean temperature. This approach basically assumes that the deep ocean is a weak damping (infinite heat sink) to interannual to decadal time scales’ natural climate variability (e.g., Alexander et al. 2000; Dommenget and Latif 2002, 2008).

However, as this deep-ocean damping will reduce the equilibrium SST response of the OZ model, it is instructive in the context of this comment to look at the SST response in a slab ocean model that does not include any deep-ocean damping. Therefore, the OZ-TLAND simulation is repeated with the OZ model replaced with a 50-m slab ocean model (hereafter SLAB–TLAND). Both simulations are integrated for 2 × 100 yr with continental surface temperatures, Tland, increased by 1 K and decreased by 1 K. The response is defined in all cases as the difference between the +1 K and the −1 K divided by two. As an analog for the OZ−2×CO2 simulation in D09, the study also simulates the slab ocean response to doubling of CO2 (referred to as SLAB−2×CO2). The near-equilibrium response in SLAB−2×CO2 is defined as the difference between the mean of the years 21–50 in the 2 × CO2 simulation minus the mean of a 50-yr control simulation.

3. Response to land surface temperature

Figure 1 shows the SST response of the OZ–TLAND and SLAB–TLAND simulations to a +1-K Tland increase. A few points regarding the response pattern can be noted here in response to the discussions in D09 and LWJ11:

  • The OZ–TLAND response pattern for the years 1–20 is not identical to that of D09, because the simulation has been started from a slightly different initial condition to produce an independent estimate to that of D09.
  • The OZ–TLAND SST response pattern is very similar across all time intervals, indicating that the pattern is a robust signature of the ECHAM5–OZ model. The amplitude is slightly increasing over time and is also tending toward a larger global mean SST warming (regions of negative SST response decrease).
  • The response pattern has some clear structure, which even includes regions that cool in the tropics and the Southern Hemisphere. This is indicating that atmospheric circulation changes are involved in the pattern formation. The structure, to some extent, resembles the hyper mode pattern of multidecadal SST variability, as discussed in Dommenget and Latif (2008).
  • It is also interesting to note that the equatorial east Pacific region tends toward a negative response to positive Tland changes, which is similar to the El Niño unrelated trends, as discussed in Compo and Sardeshmukh (2010). They argued that the trend over the last decades is unrelated to the El Niño dynamics, and is a cooling in the equatorial east Pacific region. In the context of the OZ–TLAND simulation result, the cooling found in the Compo and Sardeshmukh (2010) may be interpreted as the ocean’s response to land warming not involving El Niño dynamics.
  • The warming pattern seems also quite consistent with the Forster et al. (2000) experiment, where 3 × CO2 was increase only over land. Forster et al. also find a much weaker response over the oceans than over land, and they also found negative SST responses in the Southern Hemispheric subtropics to extratropics.
  • The SLAB–TLAND experiment shows a response pattern similar to that of the OZ-TLAND but shifted to positive values and with larger amplitudes overall. Some significant differences in the patterns reflect the different ocean dynamics of the OZ and SLAB ocean models.
The global mean response of the oceans to a +1-K Tland increase can better be discussed on the basis of the time series of global mean SST of ice-free regions, Tocean; see Fig. 2. The following points should be noted here:
  • The OZ–TLAND simulation is close to equilibrium after about 20 yr, with the equilibrium Tocean response slightly below 0.3 K, which is about 0.1 K larger than the mean of the first 20 yr. It has to be noted here that the discussion in D09 focused on the response of the ocean to the natural variability of Tland. Since the continental variability does not involve much variance on time scales longer than a year, the mean of the first 20 yr seems to be a good upper-boundary value for the ocean response. However, if the ocean’s equilibrium response to long time forcings from land is considered, as in climate change scenarios, the mean of the first 20 yr is not a sufficient estimate of the equilibrium Tocean response.
  • The SLAB–TLAND simulation is essentially in equilibrium after 20 yr, with the global mean ocean heat uptake of 0.005 ± 0.06 W m−2 in the period 51–100-yr.
  • The SLAB–TLAND response is significantly larger than in the OZ–TLAND simulation, by about 60% (0.18 K). The OZ model has a Newtonian damping to the deep ocean, which acts as a weak (~0.4 W m−2 K−1) damping to the SST, that is not present in the slab ocean model. While this damping is more realistic for interannual to decadal SST variability, it becomes unrealistic for a long time (>100 yr) equilibrium.
  • The land–sea warming response ratio (Tland/Tocean) in the OZ–TLAND experiment is 5.0 for the first 20 yr (as in D09) and 3.6 for the period 51–100 yr. The near-equilibrium SLAB–TLAND land–sea warming response ratio is 2.2 (with a ±0.1 95% confidence interval), which is still significantly larger than the 1.3 land–sea warming response ratio found in the experiment SST + 1 K in D09. Thus, clearly indicating a significant asymmetry in land–sea interaction, with the ocean forcing the land much more strongly.
Fig. 1.
Fig. 1.

SST response to Tland + 1 K forcing in different experiments.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00476.1

Fig. 2.
Fig. 2.

Tocean response to Tland + 1 K forcing in different experiments.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00476.1

4. Box model discussion

In D09 the series of different sensitivity experiments is summarized in a simple two-box model. This simple conceptual model allows for discussing how the climate system may respond in different idealized sensitivity experiments, which are, in some cases, highly hypothetical. In the following, this box model is used to discuss some of the characteristics that the climate models would have in idealized sensitivity experiments. In particular, the study will argue how the discussion may change if equilibrium responses are considered.

The box model in D09 is given by the following two equations for the tendencies of Tland and Tocean (deviations from the climatological global mean surface temperature over land and over the ice-free oceans, respectively):
e1
e2
with the feedback parameters cL and cO representing the net effects of all local feedbacks, the effective coupling parameters cLO and cOL, and the net forcings over land FL and over the ocean FO. The different heat capacities over land and ocean are given by λland and λocean, respectively. The model parameters in D09 were computed by a least squares fit to five sensitivity experiments with the OZ model. Taking the new results with the longer integration of the OZ–TLAND and the slab ocean model experiments into account, the following points can be made:
  • Given the parameters of D09, the simple box model predicts the following for the SST equilibrium response Tocean:
    e3
    This is roughly what was found in the SLAB–TLAND experiment.
  • The box model parameters can be estimated with the new experiment results: the OZ–TLAND and OZ−2×CO2 values can be replaced with those of the SLAB–TLAND and SLAB−2×CO2 equilibrium response values (Tland = 6.9 K and Tocean = 4.3 K). The FIXLAND−2×CO2 of D09 has no slab ocean counterpart, but it may be assumed, based on the SLAB–TLAND and SLAB−2×CO2 equilibrium response values, that the equilibrium ocean response of the FIXLAND−2×CO2 would be about 1.2 K. With these response values, the parameters are found to be cL = 1.6 W m−2 K−1, cO = −0.6 W m−2 K−1, cLO = 5.1 W m−2 K−1, and cOL = 0.5 W m−2 K−1, and the root-mean-square error of the fit is 0.06 K (0.09 K in D09). The parameters have not change much compared to D09, with some exception of cO, which was significantly stronger in D09, because the OZ model used in D09 does include an ocean damping term, which is not present in the slab ocean model used here.
  • We can use the box model to compare the results of D09 with the experiments 3×CO2, 3×CO2 (land only), and 3×CO2 (ocean only) of Forster et al. (2000). They find that the global warming is 29% of the 3×CO2 if only land is forced and 73% if only the ocean is forced. The box model of D09 is quite consistent with these finding, predicting 26% for land-only forcing and 74% for ocean-only forcing.

5. Summary

In summary the study discussed the response of the ocean to land forcing in sensitivity experiments in D09 in more detail. In particular, the study showed that the strength of the response of the oceans is time-scale dependent with a very weak response to land forcing on interannual time scales, as discussed in D09, and that it has about a twice as strong of a near-equilibrium response to land forcing on time scales longer than 100 yr. The asymmetric land–sea interaction, with the ocean forcing the land much more strongly than the land forces the oceans, as discussed in D09, is confirmed by this study.

Acknowledgments

I would like to thank Hugo Lambert for the discussions and for some clarification of the LWJ11 results. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) through Project DO1038/5-1 and by the Australian Research Council’s Centre of Excellence for Climate System Science.

REFERENCES

  • Alexander, M. A., , J. D. Scott, , and C. Deser, 2000: Processes that influence sea surface temperature and ocean mixed layer depth variability in a coupled model. J. Geophys. Res., 105, 16 82316 842.

    • Search Google Scholar
    • Export Citation
  • Compo, G. P., , and P. D. Sardeshmukh, 2010: Removing ENSO-related variations from the climate record. J. Climate, 23, 19571978.

  • Dommenget, D., 2009: The ocean’s role in continental climate variability and change. J. Climate, 22, 49394952.

  • Dommenget, D., , and M. Latif, 2002: Analysis of observed and simulated SST spectra in the midlatitudes. Climate Dyn., 19, 277288.

  • Dommenget, D., , and M. Latif, 2008: Generation of hyper climate modes. Geophys. Res. Lett., 35, L02706, doi:10.1029/2007GL031087.

  • Forster, P. M., , M. Blackburn, , R. Glover, , and K. P. Shine, 2000: An examination of climate sensitivity for idealised climate change experiments in an intermediate general circulation model. Climate Dyn., 16, 833849.

    • Search Google Scholar
    • Export Citation
  • Lambert, F. H., , M. J. Webb, , and M. M. Joshi, 2011: The relationship between land–ocean surface temperature contrast and radiative forcing. J. Climate, 24, 32393256.

    • Search Google Scholar
    • Export Citation
  • Roeckner, E., and Coauthors, 2003: The atmospheric general circulation model ECHAM5. Part I: Model description. MPI Rep. 349, 127 pp.

Save