Environ. Res. Lett. 10 (2015) 064011 I Takayabu et al We also examined the influence of SLR of 0.30 m, on the storm surge level with higher mean sea level condition. However, the maximum sea surface level considering a higher mean sea level condition didn’t give significant changes in the storm surge height. Theoretically, the storm surge height is inversely proportional to the depth of the bay. Since the water depth near Tacloban in Leyte Gulf has depth of approximately 10 m, 0.30 m mean sea level rise almost cancels out the change of total surge height, although there remain some dependencies on characteristics of the bay and cyclones. The combination of a storm surge and sea level rise is nonlinear, thus it is important to know how sea level rise and dynamic effects, wind and pressure surges, will change in the future. 5. Conclusion Typhoon Haiyan (local name Yolanda), the most catastrophic tropical cyclone ever to land in the western North Pacific Ocean, struck the Philippines on 8 November 2013. The typhoon and especially the storm surge in the Leyte Gulf that accompanied it killed more than 6000 people in Tacloban (Schiermeier 2013). We conducted ensemble simulations with very high resolution regional climate models and a surge model, and reproduced well the pressure depression, wind speed and surge level of Typhoon Haiyan, as an example of a worst case scenario. Furthermore, we compared these results with the results of ensemble simulations of a hypothetical natural event, one without human influences, and found that the simulated worst case typhoon and the accompanying storm surge in the real condition became worse than those in the hypothetical natural climate without anthropogenic forcing. In 15 of 16 ensemble simulations, the typhoon became stronger than it did in the hypothetical natural cases, and the height of the storm surge around Tacloban increased by around 20%. When we made the NAT external condition, we omitted anthropogenic warming in air temperature difference as well as that in SST. It is well known that vertical wind shear and mid-level entropy deficit are important in determining the intensity of tropical cyclone (Gray (1975), Tang and Emanuel (2010, 2012)). Although previous studies have suggested the weakening of Pacific Walker circulation (reductions of wind shear in the western Pacific region), and anthropogenic contributions to this observed trend, there are substantial uncertainties in observations and biases in model simulations (Tokinaga et al 2012, DiNezio et al 2013). Supplementary figure S3 shows the histograms of vertical wind shear in the 100-member ensembles of ALL-gcm and NATgcm runs of MIROC5-AGCM (Shiogama et al 2014). Although anthropogenic factors weaken the vertical wind shear in the ensemble mean (which would 8 increase the Typhoon intensity (Tang and Emanuel, 2010, 2012)), the variances due to internal variability are large. Because of the poor signal-to-noise ratio, we have not subtracted these differences in wind shear from the NAT conditions of RCM runs. It might cause our estimates of anthropogenic influence on the Typhoon intensity to become more conservative. We should also note here, that upper-ocean mixing is not accounted for because we do not have a sound assumption for prescribing the upper-ocean structure in natural climate conditions. Vincent et al (2014) insisted that the mixing has a negative feedback on TC intensity for very strong storms, but the influence of such a mechanism is not accounted in our study. As we introduced in section 1, to assess the disaster risk reduction caused by the climate change, we had quite a different approach, rather than considering the disaster mitigation as we discussed in the paper. This is to consider the disaster prevention. For this purpose, we should estimate changes in the frequency of event occurrence. However, Hansen et al (2014) suggests that it is difficult to estimate the change of likelihood with a high confidence level when the sample numbers are not high enough, especially for extreme events. To increase the sample number efficiently in cases where we are not able to conduct sufficient ensemble experiments, we have another option of introducing a stochastic hurricane model of the type used in Lin et al (2012). This method imposes a stronger limitation of the assumed hurricane structure and intensity than a dynamically downscaling approach. Thus, it is indispensable to carry out both types of approach jointly, to clarify the disaster risk of extreme events as a whole. There are uncertainties associated with estimations of the effects of anthropogenic signals on SST (Christidis and Stott 2014, Shiogama et al 2014) and attribution analysis results may also be sensitive to models used. We showed here that very high resolution regional climate models are necessary to simulate a category 5 typhoon. It should be kept in mind that more accurate representation of the typhoon with higher spatial resolution does not necessary ensure more reliable estimates of the anthropogenic contribution. Inter-comparison studies of very high resolution regional climate models should be carried out in which different estimates of anthropogenic signals are used. Acknowledgments This work was supported by the SOUSEI programmes of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. The Guiuan radar data were supplied by PAGASA, the Philippines, and Dr Kubota of Japan Agency for Marine-Earth Science and Technology (JAMSTEC) helped us to read the