The model results presented in this paper show that in the MIROC3.2 model there is a reasonably strong link between global and tropical temperature changes at the LGM and those for 2×CO2. However, there is a considerable amount of noise in the correlation, even though we are only considering the results from one model. It is clear that different processes (controlled by different parameters) affect the response to strong positive and negative forcings, even when this forcing is limited to radiative forcing of greenhouse gases. The albedo and topographical influences of large ice sheets complicate matters further, at least at the local level. Unsurprisingly, the links between regional and global scales within the same experimental epoch are much stronger. With a lack of model runs from both climate states, MD06 assumed the link existed, while Crucifix (2006) obtained results from only 4 simultaneous models and perceived no such link at the global scale. While a perturbed-parameter ensemble can form a step towards increasing our understanding, it is unlikely to cover the full range of results that structurally different models can achieve. It is therefore important, if this link is to be better understood, that directly comparable integrations of both LGM and 2×CO2 climates are performed for a larger number of GCMs in future. Furthermore, it would be helpful to ensure that the LGM boundary conditions actually represent reality as faithfully as possible, rather than representing a sensitivity analysis in which some potentially important (albeit poorly understood) elements are omitted. This is especially important if a direct comparison with data is to be attempted. Dust and vegetation changes may alter the pattern of response, and changes in ocean heat transport would increase the uncertainty of the results. Our results lend support to the idea in MD06 that the LGM Antarctic is a good place to look for a data which can be used to validate models used for climate forecasting of future GHG induced climate changes, at local, regional and global scales. Good results may in principle be obtainable using tropical temperatures, particularly those over the ocean. While the greater area in the tropics makes them an attractive area for seeking data, polar amplification of temperature changes (apparent in Fig. 2) may mean that the Antarctic provides a clearer signal relative to the uncertainties in data and model results. Our result for Greenland is not so strong, possibly due to difficulties in accurately modelling the sea ice extent.
The areas occupied by the massive northern hemisphere ice sheets and sea ice at the LGM would appear to be very poor places to seek data of relevance to GHG forcing. Our results indicate that the temperature changes in those regions are controlled by different parameters for both LGM and for the southern sea ice region for the 2×CO2 climate. This implies different processes at work in those regions which therefore means that changes observed at the present day in the southern sea ice locations would provide only relatively weak information on the value of future globally averaged warming.
The MIROC3.2 model shows an asymmetry in climate sensitivity calculated by decreasing rather than increasing the greenhouse gases, with 80% of the ensemble having a weaker cooling than warming. This asymmetry, if confirmed by other studies, would mean that direct estimates of climate sensitivity from the LGM are likely to be underestimated by the order of half a degree. However, this result may be model dependent. Analysis of the parameters varied in the model suggest the asymmetrical response may be linked to the behaviour of ice in the clouds, which is therefore indicated as an important area for future research.
Acknowledgements. We thank the K-1 Japan project members for support and discussion, and in particular Tokuta Yokohata for the radiative forcing calculations. This work was partially supported by the Research Revolution 2002 (RR2002) of the Ministry of Education, Sports, Culture, Science and Technology of Japan. The model calculations were made on the Earth Simulator of JAMSTEC. The research was partially supported by JSPS Grant-in-Aid for Scientific Research (B) no. 16340136.
Answers to all your Biology Questions
