SUPPLEMENTARY INFORMATIONdoi: 10.1038/ngeo779nature geoscience | www.nature.com/naturegeoscience 11Progress Article:Tropical Cyclones and Climate ChangeSupplemental MaterialFebruary5, 2010ThomasR.Knutson*1, John L.McBride2, Johnny Chan3, Kerry Emanuel4, Greg Holland5, Chris Landsea6, Isaac Held1, James P.Kossin7, A. K. Srivastava8, and MasatoSugi91Geophysical Fluid Dynamics Laboratory/NOAA, 201 Forrestal Road, Princeton, NJ08542, USA2Centre for Australian Weather and Climate Research, Melbourne, Australia 30013Guy Carpenter Asia-Pacific Climate Impact Centre, City University of Hong Kong,Kowloon, Hong Kong, China4Program in Atmospheres, Oceans, and Climate, Massachusetts Institute of Technology, Rm. 54-1620 MIT, 77 Massachusetts Avenue, Cambridge, MA 02139, USA.5National Center for Atmospheric Research, Boulder, CO, USA.6National Hurricane Center/NWS/NOAA, 11691 SW 17thStreet, Miami, FL, 33165, USA7National Climatic Data Center/NOAA, 1225 W. Dayton St., Madison, WI, 53706, USA8India Meteorological Department, Shivajinagar, Pune 411005, India9Research Institute for Global Change, JAMSTEC, 3173-25 Showa-machi, Kanazawa-ku, Yokohama, 236-0001 Kanagawa, JapanContact Information:Thomas R. KnutsonGFDL/NOAA201 Forrestal RoadPrinceton, New Jersey 08542 U.S.A.Ph: +1-609-584-7152Fax: +1-609-987-5063Email: [email protected] nature geoscience | www.nature.com/naturegeoscienceSUPPLEMENTARY INFORMATIONdoi: 10.1038/ngeo7792Contents of Supplemental Material:S1. Review of tropical cyclonefrequency projections. S2. Review of intensity projectionsS3. Review of rainfall projectionsS4. Some definitionsS5. Limitations of tropical cyclone historical dataS6.Comparison with Previous AssessmentsS7. Recommendations for Future ProgressAdditional References for Supplemental MaterialS1. Review of tropical cyclone frequency projections.Projections of tropical cyclone frequency changes for various climate warming scenarios have been made by counting tropical cyclone-like vortices in regional or global climate models. Table S1 summarizes the tropical cyclone frequency projections from 17modeling studies using models having horizontal grid spacing equivalent to about 120 km or less. The top part of the table summarizes changes for storms of tropical storm intensity or greater. Globally, all of the models show a decrease in this metric, ranging from-6% to -34%, though not statistically significant in at least one case. A consistent sign of change (decrease) is also seen for the Southern Hemisphere mean frequency. Fornature geoscience | www.nature.com/naturegeoscience 3SUPPLEMENTARY INFORMATIONdoi: 10.1038/ngeo7792Contents of Supplemental Material:S1. Review of tropical cyclonefrequency projections. S2. Review of intensity projectionsS3. Review of rainfall projectionsS4. Some definitionsS5. Limitations of tropical cyclone historical dataS6.Comparison with Previous AssessmentsS7. Recommendations for Future ProgressAdditional References for Supplemental MaterialS1. Review of tropical cyclone frequency projections.Projections of tropical cyclone frequency changes for various climate warming scenarios have been made by counting tropical cyclone-like vortices in regional or global climate models. Table S1 summarizes the tropical cyclone frequency projections from 17modeling studies using models having horizontal grid spacing equivalent to about 120 km or less. The top part of the table summarizes changes for storms of tropical storm intensity or greater. Globally, all of the models show a decrease in this metric, ranging from-6% to -34%, though not statistically significant in at least one case. A consistent sign of change (decrease) is also seen for the Southern Hemisphere mean frequency. For 3the Northern Hemisphere mean, several models show a decrease, but a few models indicate essentially no change. At the individual basin scale, especially for the Northern Hemisphere basins, the sign of the projected changes is much more variable across different models. For example, for the North Atlantic basin, 10modelexperimentsproject an increase, and 13 a decrease. In the Northeast Pacific, three models report an increase, three models a decrease, and one essentially no change. The magnitude of the projected changes for individual basins in these studies ranges up to +/-50% or more.The bottom section of Table S1 presents some results for tropical cyclone frequency, considering storms of higher intensities than the minimum tropical storm intensity threshold. Eight studies report increases in the frequency of higher intensity tropical cyclones, although the specific intensity threshold considered varies, since the various models simulate storms only to certain intensity thresholds due in large part to their limited resolution. While three studies reported no change of frequency of storms at any intensity simulated, and one study (for a single basin) reported a decrease of frequency at all intensities, all of these latter four studies were conducted with relatively low resolution models (~120 km grid). We regard such lower resolution models as being less credible for simulating higher intensity categories of tropical cyclones.To summarize the models’ frequency projections, a robust decrease in tropical cyclone frequency is projected globally (-6% to -34%) and for the southern hemisphere, while the projections are much more mixed (and the uncertainty of projections is much larger) for the individual basins (projected changes up to +/-50% or more). In some cases these4 nature geoscience | www.nature.com/naturegeoscienceSUPPLEMENTARY INFORMATIONdoi: 10.1038/ngeo7794larger changes projected for the individual basins by downscaling or testingsingle modelrealizations10,11,45maybe exaggerating the model’s actual response to climate forcings, due to unintended influences of internal climate variability on the projections.There is a clear tendency in the models, particularly at higher resolution (60 km grid or less), to project anincrease in the frequency of the stronger classes of tropical cyclones, although the actual intensity class of these strong model cyclones varies depending on the various limitations of each model. (e.g., resolution, physics).S2. Review of intensity projectionsA key issue for future projections of tropical cyclone intensity is the realism of the model or theory used in projecting the intensities of the cyclones. As of the 2008 season, the minimum horizontal grid