Climatic Change The potential impacts of 21st century climatic and population changes on human exposure to the virus vector mosquito Aedes aegypti Manuscript Number: Full Title: Article Type: Corresponding Author: Corresponding Author Secondary Information: Corresponding Author?s Institution: Corresponding Author?s Secondary Institution: First Author: First Author Secondary Information: Order of Authors: Order of Authors Secondary Information: Funding Information: Abstract: --Manuscript Draft-- The potential impacts of 21st century climatic and population changes on human exposure to the virus vector mosquito Aedes aegypti O'Neill Special Issue Andrew Monaghan National Center for Atmospheric Research UNITED STATES National Center for Atmospheric Research Andrew Monaghan Andrew Monaghan Kevin Sampson Daniel Steinhoff Kacey Ernst Kristie Ebi Bryan Jones Mary Hayden National Science Foundation National Institutes of Health (IR01AI091843) Dr. Andrew Monaghan Kacey Ernst The mosquito Aedes aegypti transmits the viruses that cause dengue and chikungunya, two globally?important vector?borne diseases. We investigate how choosing alternate emissions and/or socioeconomic pathways may modulate future human exposure to Ae. aegypti. Occurrence patterns for Ae. aegypti for 2061?2080 are mapped globally using empirically downscaled air temperature and precipitation projections from the Community Earth System Model, for the Representative Concentration Pathway (RCP) 4.5 and 8.5 scenarios. Population growth is quantified using gridded global population projections consistent with two Shared Socioeconomic Pathways (SSPs), SSP3 and SSP5. Change scenarios are compared to a 1950?2000 reference period. A global land area of 56.9 km"2 is climatically suitable for Ae. aegypti during the reference period, and is projected to increase by 8% (RCP4.5) to 13% (RCP8.5) by 2061?2080. The annual average number of people exposed globally to Ae. aegypti for the reference period is 3794 M, a value projected to statistically significantly increase by 298?460 by 2061?2080 if only climate change is considered, and by 4805?5084 (127?134%) for SSP3 and 2232?2483 (59?65%) for SSP5 considering both climate and population change (lower and upper values of each range represent RCP4.5 and RCP8.5 respectively). Thus, taking the lower?emissions RCP4.5 pathway instead of RCP8.5 may mitigate future human exposure to Ae. aegypti globally, but the effect of population growth on exposure will likely be larger. Regionally, Australia, Europe and North America are projected to have the largest Powered by Editorial Manager? and ProduXion Manager? from Aries Systems Corporation percentage increases in human exposure to Ae. aegypti considering only climate change. Response to Reviewers: See Attachment Powered by Editorial Manager? and ProduXion Manager? from Aries Systems Corporation Manuscript Click here to download Manuscript monaghan_etal_brace_aedes_paper_text_v5.docx Click here to view linked References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 1 The potential impacts of 21st century climatic and population changes on human 2 exposure to the virus vector mosquito Aedes aegypti 3 A.J. Monaghan, K.M. Sampson, D.F. Steinhoff, K.C. Ernst, K.L. Ebi, B. Jones and M.H. 4 Hayden 5 6 Andrew J. Monaghan – Corresponding Author 7 National Center for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307 8 Phone: 303-497-8424 9 Fax: 303-497-8401 10 monaghan@ucar.edu 11 12 M.H. Hayden, K.M. Sampson, D.F. Steinhoff 13 14 National Center for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307 15 K.C. Ernst 16 University of Arizona, College of Public Health, P.O. Box 245163, Tucson, AZ 85724 17 18 K.L. Ebi 19 20 University of Washington, School of Public Health, 1705 NE Pacific St, Box 357965, Seattle, WA 98195-7965 21 22 B. Jones 23 24 25 City University of New York, CUNY Institute for Demographic Research, 135 East 22nd St, New York, NY 10010 26 27 Submitted to Climatic Change as part of a Special Issue on the Benefits of Reduced Anthropogenic Climate changE (BRACE) 28 29 30 Keywords: Climate change, Aedes aegypti, dengue, chikungunya, BRACE 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 31 32 Abstract The mosquito Aedes (Ae). aegypti transmits the viruses that cause dengue and 33 chikungunya, two globally-important vector-borne diseases. We investigate how choosing 34 alternate emissions and/or socioeconomic pathways may modulate future human exposure to 35 Ae. aegypti. Occurrence patterns for Ae. aegypti for 2061-2080 are mapped globally using 36 empirically downscaled air temperature and precipitation projections from the Community 37 Earth System Model, for the Representative Concentration Pathway (RCP) 4.5 and 8.5 38 scenarios. Population growth is quantified using gridded global population projections 39 consistent with two Shared Socioeconomic Pathways (SSPs), SSP3 and SSP5. Change 40 scenarios are compared to a 1950-2000 reference period. A global land area of 56.9 M km2 is 41 climatically suitable for Ae. aegypti during the reference period, and is projected to increase 42 by 8% (RCP4.5) to 13% (RCP8.5) by 2061-2080. The annual average number of people 43 exposed globally to Ae. aegypti for the reference period is 3794 M, a value projected to 44 statistically significantly increase by 298-460 M (8-12%) by 2061-2080 if only climate 45 change is considered, and by 4805-5084 M (127-134%) for SSP3 and 2232-2483 M (59- 46 65%) for SSP5 considering both climate and population change (lower and upper values of 47 each range represent RCP4.5 and RCP8.5 respectively). Thus, taking the lower-emissions 48 RCP4.5 pathway instead of RCP8.5 may mitigate future human exposure to Ae. aegypti 49 globally, but the effect of population growth on exposure will likely be larger. Regionally, 50 Australia, Europe and North America are projected to have the largest percentage increases in 51 human exposure to Ae. aegypti considering only climate change. 52 2 53 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 54 1. Introduction The mosquito Aedes (Ae). aegypti serves as the primary vector for the viruses that cause 55 dengue, the most prevalent vector-borne viral disease (WHO 2009). About 2.5 billion people 56 in over 100 countries live in regions of high dengue risk (Kroeger and Nathan 2005), and a 57 recent study estimated that up to 390 million dengue virus infections occur worldwide 58 annually (Bhatt et al. 2013). When present, the febrile symptoms of dengue virus infections 59 are generally mild to moderate, but about 1% of infections are manifested as the more severe 60 dengue hemorrhagic fever (Gubler 1998). Aedes aegypti is also a vector of chikungunya 61 virus, which can cause dengue-like symptoms in infected persons (Pialoux et al. 2007). The 62 annual global occurrence of chikungunya is not well known; however, sporadic epidemics 63 affecting >1M people have occurred in Asia and, more recently, the Americas (Pialoux et al. 64 2007, Nasci 2014, Halstead 2015). The ranges of dengue and chikungunya are expanding 65 (Eisen and Moore 2013, Morrison 2014), making Ae. aegypti a vector of growing importance. 66 Climate variability largely determines the distribution and population dynamics of Ae. 67 aegypti at the global scale (e.g., Capinha et al. 2014). Aedes aegypti is typically found in 68 tropical and subtropical regions worldwide, within urban areas where it can exploit artificial 69 water-filled containers for its immature (larval and pupal) stages (Tabachnick and Powell 70 1979). Its northern (southern) hemisphere range is generally equatorward of the January 71 (July) average 10ºC isotherm, but in some cases can extend as far poleward as the 1.8ºC 72 January average isotherm in North America and Europe (Christophers 1960). At local scales, 73 temperature, humidity and precipitation influence Ae. aegypti development, reproduction and 74 survival (Focks et al. 1993, Morin et al. 2013). Warmer temperatures are associated with 75 higher adult mosquito abundance, because immature Ae. aegypti development rates increase 76 as water temperatures rise concurrently with air temperatures (Eisen et al. 2014). However, 77 extremely high temperatures (above ~ 32oC) can decrease development rates and increase 3 78 1 2 79 3 4 5 80 6 7 8 81 9 10 82 11 12 13 83 14 15 84 16 17 85 18 19 20 86 21 22 87 23 24 25 88 26 27 89 28 29 30 90 31 32 91 33 34 35 92 36 37 93 38 39 94 40 41 42 95 43 44 96 45 46 47 97 48 49 50 98 51 52 99 53 54 55 100 56 57 101 58 59 102 60 61 62 63 64 65 adult mortality (Christophers 1960). Higher rainfall generally increases the number of suitable container habitats for Ae.aegypti, and higher humidity prevents desiccation of adults, facilitating greater mosquito abundance (Lucio et al. 2013). Numerous studies employing varying methodologies have examined the potential for anthropogenic climate change to alter the global range of Ae. aegypti (Campbell et al. 2014, Capinha et al. 2014, Khormi and Kumar 2014, Rogers 2015). Collectively, the results suggest that climate change (in particular warmer temperatures) may cause the future range of Ae. aegypti to expand poleward from its present margins of survival. However, within some regions presently suitable for Ae. aegypti, the future results are not as consistent among studies. Less suitable habitats may result from climatic conditions that become too hot or dry (Rogers 2015). The effects of rising temperatures on Ae. aegypti populations are comparatively robust among the studies that consider temperature as a variable, but the effects of precipitation shifts are less certain. Manual filling of mosquito container habitats by humans may dampen the response of Ae. aegypti to rainfall fluctuations (Kearney et al. 2009), and in some instances, increased water storage practices in response to locally drier conditions may even enhance future habitat suitability for Ae. aegypti (Beebe et al. 2009). Some argue that such human factors, and others including intervention practices (e.g., pesticide use), urbanization, and transportation networks, may be stronger determinants of the range of Ae. aegypti than climate (e.g., Gubler 2002, Parham et al. 2015). This paper is part of a project on the Benefits of Reduced Anthropogenic Climate changE (BRACE;;  O’Neill  and  Gettelman 2015.). BRACE aims to characterize the differential climate outcomes and subsequent impacts on a variety of sectors that result from differences in greenhouse gas (GHG) forcing from two distinct emissions pathways: one in which GHG emissions increase throughout the 21st century, and the other in which GHG emissions eventually decrease. Related BRACE papers on health impacts examine population exposure 4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 103 to extreme heat events (Jones et al. 2015, Xu et al. 2015) and future risks of heat-related 104 mortality (Anderson et al. 2015, Marsha et al. 2015). Here, we define four types of 105 occurrence patterns for the virus vector mosquito Ae. aegypti on the basis of temperature and 106 precipitation thresholds following Eisen et al. (2014). An ensemble of global climate model 107 simulations is then employed to characterize how these patterns may shift in the late 21st 108 century for two different climate change outcomes associated with the two distinct GHG 109 emissions pathways. Finally, two plausible population pathways for humanity are combined 110 with the projected mosquito occurrence patterns to estimate the number of humans exposed 111 to Ae. aegypti in the late 21st century. The present study is unique in that 1) it uses 112 straightforward climatic thresholds to define distinct Ae. aegypti occurrence patterns; 2) it 113 explores the relative roles of climatic versus population changes on the future number of 114 people exposed to Ae. aegypti; and 3) it quantifies the avoided climatic impacts of following 115 a lower versus higher intensity GHG emissions pathway. 116 2. Methods 117 A brief overview of methods is provided here. A full description is in Online Resource 1. 118 Eisen et al. (2014) establish four types of occurrence patterns for Ae. aegypti, based on 119 temperature and precipitation thresholds in representative cities in which Ae. aegypti is 120 present: 121 122 123 124 125 126 Type 1. Year-around potential for high abundance of Ae. aegypti (e.g., Manila, Philippines); Type 2. Year-around presence of Ae. aegypti but only seasonal potential for high abundance (e.g., Chiang Mai, Thailand); Type 3. Only seasonal presence of Ae. aegypti active stages (larvae, pupae and adults), with overwintering eggs (e.g., Buenos Aires, Argentina); 5 127 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 128 Type 4. Only seasonal presence of Ae. aegypti active stages, with no overwintering eggs producing viable larvae (e.g., Puebla City, Mexico). 129 These distinctions are important, because they determine the duration and "intensity" to 130 which humans in a given region may be exposed to Ae. aegypti. Table 1 shows the 131 temperature and precipitation thresholds from Eisen et al. (2014) used here to define Types 1- 132 4. Only one type is allowed to occur per grid box; where types overlap, the more suitable 133 type for Ae. aegypti is selected (e.g., if Types 1 and 2 overlap, Type 1 is selected). 134 Collectively, these thresholds indicate that warmer, wetter regions with small temperature 135 variability are best suited for Ae. aegypti. The thresholds from Table 1 are applied to the 136 gridded global temperature and precipitation fields described below to produce reference and 137 future maps of the four Ae. aegypti occurrence patterns. 138 A gridded 10 arc-minute monthly global climatology of near-surface temperature and 139 precipitation for 1950-2000 was obtained from version 1.4 of WorldClim (Hijmans et al. 140 2005). WorldClim comprises the reference historical data used to produce maps of Ae. 141 aegypti occurrence patterns, and provides the basis for a simple procedure to empirically 142 downscale and bias correct projected climate model data described in Online Resource 1. 143 Maps of Ae. aegypti occurrence patterns for the late 21st century are produced with future 144 climate projections from version 1 of the Community Earth System Model (CESM; Hurrell et 145 al. 2013). CESM is a fully coupled atmosphere-ocean global climate model (AOGCM). The 146 version used here is the same that supported the Coupled Model Intercomparison Experiment 147 Phase 5 (CMIP5; Taylor et al. 2012). Simulations from the historical period (1950-2000) and 148 two future projections for 2061-2080 are used. The future simulations are forced in 149 accordance with the Representative Concentration Pathway (RCP; van Vuuren et al. 2011) 150 4.5 and 8.5 scenarios. RCP4.5 is a low-to-moderate emissions and concentration scenario 151 with GHG radiative forcing reaching 4.5 W m-2 near 2100. RCP8.5 is a high-emissions and 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 152 concentration scenario with GHG radiative forcing reaching 8.5 W m-2 near 2100. Monthly 153 near-surface air temperature and precipitation fields are used from the first 15 members of a 154 30-member CESM initial condition ensemble based on the RCP8.5 scenario (Kay et al. 2014) 155 and from a 15-member CESM ensemble based on the RCP 4.5 scenario (Sanderson et al. 156 2015), allowing internal model variability to be characterized. The differences in CESM 157 temperature and precipitation between 1950-2000 and 2061-2080 for each ensemble member 158 are then added to the corresponding WorldClim fields to generate bias-corrected, downscaled 159 future projections, as described in Online Resource 1. 160 To assess the potential impacts of future population changes on exposure to Ae. aegypti, 161 gridded global population projections (Jones and O'Neill 2015) from two of the five Shared 162 Socioeconomic Pathways (SSPs) – SSP3 and SSP5 – are used.  The  SSP3,  or  “Regional 163 Rivalry”,  pathway describes a world in which there are high levels of inequality and little 164 coordination across countries, and in which little progress is made toward meeting 165 development goals, addressing environmental issues, or reducing fossil fuel consumption 166 (O'Neill et al. 2015). Slow economic development and lower education trends stall the 167 demographic transition in lower income countries, leading to rapid population growth relative 168 to higher income nations, and a steadily increasing global population throughout the century. 169 The  SSP5,  or  “Fossil-fueled development”, pathway describes a world in which traditional 170 economic growth is seen as the solution to social and economic issues, and is characterized 171 by greater progress toward human development, but at the cost of intense usage of resources 172 and high fossil fuel consumption (O'Neill et al. 2015). Economic growth is rapid and global 173 population peaks by mid-century and then declines, driven by declining fertility in 174 developing countries, while urbanization rates are very high worldwide. The gridded SSP3 175 and SSP5 population projections span each decade from 2010 to 2100, and are based on 176 version 3 of the Gridded Population of the World (GPW) dataset for the year 2000 (CIESIN, 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 177 2005), and aggregate country-level population (KC and Lutz 2014) and urbanization (Jiang 178 and  O’Neill  2015)  projections  to  2100  consistent  with  each  SSP. The SSP3 and SSP5 179 projections for 2070 are used to represent the 2061-2080 period, and the GPW 2000 dataset 180 to represent the 1950-2000 reference period. The gridded population datasets are 181 conservatively remapped from their native 7.5 arc-minute spatial resolution to the same 10 182 arc-minute resolution as the climate fields, and are used to estimate the recent and future 183 number of persons exposed to Ae. aegypti for each occurrence pattern (Types 1-4), by 184 summing the population for all grid cells corresponding to a given pattern and region of 185 interest. Regions include the entire globe and the six permanently inhabited continents: 186 Africa, Asia, Australia (including Oceania), Europe, North America and South America. 187 In all results presented below, uncertainty due to both the mapping methodology and 188 internal climate model variability is considered. A validation of the global reference maps of 189 Ae. aegypti occurrence patterns and the number of humans exposed is described in Online 190 Resource 1, yielding an estimated methodological uncertainty of +/-10%. Statistical 191 significance between means is assumed when a two-tailed Student's t-test is p<0.05. 192 3. Results 193 3.1 Future changes in climate fields 194 Widespread and statistically significant near-surface air temperature increases over land 195 are projected for the RCP4.5 and RCP8.5 scenarios for 2061-2080, consistent with rising 196 greenhouse gases in both emissions pathways (See Online Resource 2, Fig. S1). The global 197 average near-surface temperature increase (over land) for 2061-2080 relative to 1950-2000 is 198 3.1 oC for RCP4.5 and 4.8 oC for RCP8.5. The largest warming is typically at middle to high 199 latitudes. Statistically significant precipitation changes for 2061-2080 over land are also 200 widespread for both RCP scenarios. Most regions exhibit increased precipitation, consistent 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 201 with an overall increase in rainfall globally due to greater atmospheric water vapor resulting 202 from warmer tropospheric temperatures (e.g., Held and Soden 2006). Notable exceptions 203 where less rainfall is projected include Amazonian South America, Central America, western 204 Africa, the Iberian Peninsula, southern Africa, northwestern India and Pakistan, and southern 205 Australia. These patterns of climatic change play an important role in the projected 2061- 206 2080 changes in the occurrence patterns of Ae. aegypti. 207 3.2 Future changes in the geographic distribution of Ae. aegypti 208 The global extents of the four Ae. aegypti occurrence patterns for the 1950-2000 reference 209 period and their changes for the 2061-2080 RCP scenarios are shown in Fig. 1. In general, 210 the most suitable occurrence patterns (Types 1 and 2) are near the equator and the less 211 suitable patterns (Types 3 and 4) are poleward (Fig. 1a), following the climatic gradients 212 from the relatively warm, wet tropics to the cooler, drier mid-latitudes. Table 2 accompanies 213 Fig. 1 and presents the total surface area for the 1950-2000 reference period and the change 214 in surface area for the two 2061-2080 RCP scenarios, for each continent and globally. Types 215 1 and 2 are projected to expand on all continents for both the RCP4.5 and RCP8.5 scenarios, 216 with the exception of Europe where conditions remain too cool. Interestingly, strong drying 217 in the Amazon rainforests for RCP8.5 leads to a contraction of Type 1 in that area (Fig. 1b), 218 muting the overall expansion of Type 1 in South America. Type 3 exhibits areas of both 219 expansion and contraction, depending on the continent. Statistically significant contraction 220 of Type 3 occurs in Africa, Asia, Australia (RCP8.5 only), South America and globally, and 221 is mainly due to the encroachment of the more suitable Type 2 pattern into areas previously 222 categorized as Type 3. Significant northward expansion of Type 3 occurs in Europe and 223 North America into areas previously categorized as Type 4. Type 4 significantly expands 224 poleward into previously unsuitable areas in Asia, Europe and North America due to warmer 225 temperatures, and expands in mainland Australia (where temperatures are already suitable) 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 226 due to increases in rainfall. Type 4 significantly contracts in Africa (RCP4.5 only) and South 227 America due to replacement by more suitable types. Globally there is a statistically 228 significant expansion of Types 1, 2, 4, and "All Types" combined, and a contraction of Type 229 3. On a percentage basis, the "All Types" expansion is 8-13% globally depending on the 230 RCP scenario, influenced by particularly large expansion of Type 1 (44-54%) and Type 2 231 (15-33%). In most cases where expansion occurs, it is larger under the RCP8.5 scenario 232 compared to RCP4.5. 233 3.3 Future changes in human exposure to Ae. aegypti 234 Figure 2 shows the projected number of persons exposed to Ae. aegypti globally for 2061- 235 2080 as a result of changes to the areal extents of the four occurrence patterns. Similar figures 236 for each of the 6 continents are in Online Resource 2. Table 3 accompanies Figure 2 and 237 presents the changes numerically for "All Types" combined as well as for the difference 238 between RCP8.5 and RCP4.5, globally and for each continent. The GPW pathway assumes 239 that population does not change from year 2000 levels, and therefore isolates the specific 240 impact of climate change on persons exposed to Ae. aegypti. The GPW results for 2061-2080 241 are largely consistent with the changes in areal extent discussed in the previous section; 242 globally the number of persons exposed increases for Types 1, 2, 4 and "All Types" and 243 decreases for Type 3. When population growth is accounted for in addition to climate change 244 (the SSP3 and SSP5 pathways), in nearly every instance changes in the number of persons 245 exposed to Ae. aegypti far exceed those due to climate change alone (GPW). Globally, the 246 number of persons exposed for the 1950-2000 reference period for "All Types" is 3794 247 million; this value expands by 298 million (8%) for the GPW case (climate change only), but 248 by 4805 million (127%) and 2232 million (59%) for the SSP3 and SSP5 pathways. 249 250 Changes in the number of persons exposed under SSP3 are larger than SSP5 for continents primarily comprised of developing and middle-income countries (Africa, Asia and 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 251 South America) because population grows most rapidly in developing countries for the SSP3 252 pathway. By contrast, for continents with predominantly industrialized countries (Australia, 253 Europe and North America), changes are largest for SSP5 because population grows most 254 rapidly in industrialized countries for this pathway. On a percentage basis, growth of persons 255 exposed for the climate change only pathway (GPW) is greatest for the continents with a 256 large fraction of land in the mid-latitudes: Australia, Europe and North America. For 257 example, Europe exhibits particularly large percentage growth in persons exposed to Ae. 258 aegypti for the GPW pathway of 144% for RCP4.5 and 292% for RCP8.5, and trails only 259 much-larger Asia in terms of absolute growth. 260 Compared to RCP4.5, there is larger growth in the number of persons exposed under the 261 RCP8.5 scenario on all continents and for all 3 population pathways for "All Types" (Table 262 3). However, only for North America and Europe are the differences significant for all three 263 population pathways. The differences are insignificant for all three of the population 264 pathways for Africa, Asia and South America, where strong population growth far outweighs 265 the impact of climate change. For RCP8.5 an additional 162 (GPW), 279 (SSP3) and 251 266 (SSP5) million persons globally would be exposed to Ae. aegypti compared to RCP4.5, 267 equivalent to about 4-7% of the 3974 million people exposed for the 1950-2000 reference 268 period, though only the SSP5 changes are significant. 269 4. Discussion and Conclusions 270 By 2061-2080, for both RCP scenarios, many regions already suitable for Ae. aegypti 271 may maintain or shift toward Type 1 and Type 2 occurrence patterns associated with greater 272 mosquito abundance for a longer annual duration. The areal extent of the Type 1 pattern is 273 projected to increase by 44-54% globally, and Type 2 by 15-33%, with the majority of 274 expansion in Africa, southern Asia and South America. Accordingly, the number of persons 275 exposed to these more suitable Ae. aegypti occurrence patterns may increase substantially. 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 276 For example, the number of humans exposed to Ae. aegypti in 2061-2080 for the Type 1 277 pattern is projected to increase by about 35-45% (GPW), 200-215% (SSP3) and 115-120% 278 (SSP5) from a present-day exposed population of about 230 million. Percentage increases 279 are similar for the Type 2 pattern. Comparing these results to the global "All Types" increases 280 in humans exposed by 2061-2080 of 8-12% (GPW), 127-134% (SSP3), and 59-64% (SSP5), 281 it is evident that even climate change alone (the GPW case), in the absence of population 282 growth, leads to relatively large increases of these more suitable occurrence patterns. 283 In regions that have marginal or no suitability in the present-day, by 2061-2080 a shift 284 toward enhanced seasonal suitability for Ae. aegypti is projected, particularly due to the 285 expansion of the Type 4 occurrence pattern into areas previously unsuitable for the 286 establishment of Ae. aegypti. The areal extent of the Type 4 pattern is projected to grow by 287 8-18% globally, but will likely grow at a more rapid rate at the cool mid-latitude margins of 288 present-day suitability, particularly in Europe (138-308%). Consequently, the number of 289 Europeans exposed to Ae. aegypti in 2061-2080 for the Type 4 pattern is projected to increase 290 by about 140-265% (GPW), 105-220% (SSP3) and 240-400% (SSP5) from a present-day 291 exposed population of 44 million. Therefore, in addition to population change, by 2061- 292 2080 climate change alone (the GPW case) may drive large increases in the number of 293 humans exposed seasonally to Ae. aegypti in areas that have marginal present-day suitability. 294 We therefore find that within regions, for certain Ae. aegypti occurrence patterns, climate 295 change plays an important role in driving increases in human exposure to Ae. aegypti. 296 However, when considering human exposure to Ae. aegypti for all occurrence patterns 297 collectively, we find that climate change is a comparatively modest driver of increased 298 exposure relative to population growth globally and particularly for continents comprised 299 mainly of developing and middle income countries. For example, the average growth in the 300 population exposed to all Ae. aegypti occurrence patterns by 2061-2080 for Africa, Asia and 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 301 South America combined is 4-6% for GPW (climate change only), compared to 161-165% 302 for SSP3 and 70-72% for SSP5 (pathways for which both climate change and population 303 growth are accounted for). As a result, the differences in human exposure to Ae. aegypti 304 between the RCP4.5 and RCP8.5 climate change scenarios are statistically insignificant for 305 all three continents for all three population pathways, meaning there is no apparent advantage 306 to taking the lower-emissions RCP4.5 pathway in terms of reducing human exposure to Ae. 307 aegypti in these regions. There are however, statistically significant differences between 308 RCP4.5 and RCP8.5 for North America and Europe for all three population pathways, and 309 globally for the SSP5 storyline, and thus in these regions there is an advantage to taking the 310 lower emissions RCP4.5 pathway, though it is still modest compared to overall changes in 311 exposure due to population growth. 312 Total population growth in developing and middle income countries in Africa, Asia and 313 South America is greater for the SSP3 ("Regional Rivalry") pathway compared to the SSP5 314 ("Fossil-fueled Development") pathway, due to more rapidly declining birthrates in 315 developing and middle income countries for SSP5 (O'Neill et al. 2015). Conversely, 316 population growth is larger in developed countries in Australia, Europe and North America 317 for SSP5, driven by urbanization. As a result, human exposure to Ae. aegypti increases most 318 for the SSP3 pathway in Africa, Asia and South America, whereas it increases most for the 319 SSP5 pathway in Australia and Europe (and is about the same for either pathway in North 320 America due to developing and middle income countries in Central America versus the 321 developed United States). These are important distinctions because SSP3 and SSP5 also 322 differ in dimensions not directly included in our modeling results, but that would be expected 323 to affect the extent of future outbreaks of dengue and their control. SSP3 is a world with 324 weak investments in health, institutions, and research and development with a declining 325 ability to identify, communicate, prevent and manage dengue outbreaks (Ebi 2013). Most 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 326 urban growth in low-income countries is in unplanned settlements, with limited attention to 327 increasing access to safe water and improved sanitation, implying increases in the number of 328 possible breeding sites for Ae. aegypti. On the other hand, SSP5 is a world with strong 329 investments in health, education, institutions, and research and development with the capacity 330 and political will to prepare for, prevent and manage dengue outbreaks (Ebi 2013), even 331 though enhanced global trade networks (and presumably transportation infrastructure) may 332 facilitate the expansion and/or reintroduction of Ae. aegypti into marginally suitable areas. 333 Significant progress is achieved in improving access to safe water and improved sanitation, 334 thus reducing breeding grounds for Ae. aegypti. These distinctions imply that the more 335 vulnerable developing and middle income countries in Africa, Asia and South America may 336 be disproportionately impacted under the SSP3 pathway, not only because more humans in 337 these countries would be exposed to Ae. aegypti under SSP3, but also because of 338 comparatively diminished capacity to mitigate dengue risk. 339 The analysis is subject to a number of model and data limitations. We employ a simple 340 modeling approach for Ae. aegypti occurrence based on long-term monthly and annual 341 average temperature and precipitation thresholds. This approach does not directly address 342 important complex system interactions related to Ae. aegypti ecology, development, behavior 343 and survival that occur on shorter timescales, and that more sophisticated process-based 344 modeling approaches may resolve (e.g., Focks et al. 1993, Morin et al. 2015). For example, 345 several climatic processes known to impact Ae. aegypti life-history traits are not represented, 346 such as rainfall intensity (Koenraadt et al. 2008), diurnal temperature variability (Carrington 347 et al. 2013) and upper temperature limits (Williams et al. 2014). Our approach also neglects 348 microclimatic impacts that may buffer mosquitoes from temperature and precipitation 349 extremes (Kearney et al. 2009). It is noteworthy that while the projections of populations 350 exposed to Ae. aegypti (Fig. 2 and Table 3) apply only to areas that are inhabited by humans, 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 351 our calculations of the areal extent of occurrence patterns (Fig. 1 and Table 2) are 352 independent of population and thus imply suitability in some uninhabited regions. Only 353 results from the CESM AOGCM are used, an approach that does not allow inter-model 354 uncertainty to be estimated (CESM is one of about 20 CMIP5 AOGCMs). The SSP 355 population projections are subject to numerous sources of uncertainty, not the least the 356 unknown geopolitical landscape of the future (O'Neill et al. 2015). 357 Other study limitations are related to human and mosquito behavior. Aedes aegypti is 358 highly dependent on human behavior for finding suitable container habitats, shade, and 359 shelter (e.g., Eisen et al. 2014). For example, a study in San Juan, Puerto Rico found that 360 most Ae. aegypti pupae were produced in human-managed containers (Barrera et al. 2011). 361 How such behavioral factors may change in the future is unaccounted for. Interspecies 362 competition among Ae. aegypti and other mosquitoes may affect the rate and extent of range 363 expansion; for example, it has been suggested that Ae. albopictus may reduce or displace Ae. 364 aegypti populations in regions of the United States, Brazil and Thailand where both species 365 co-occur (Moji et al. 1988, Lounibos et al. 2002, Braks et al. 2004). How human 366 interventions aimed at reducing Ae. aegypti populations may change in the future is 367 unknown. For example, controversial releases of genetically-modified "sterile" male 368 mosquitoes (Ernst et al. 2015) may become more common in the future, and, if they do, 369 would differ between the SSPs. Additionally, how other human factors such as cultural 370 practices, water access, urbanization, transportation networks and global trade may evolve 371 and impact the spread of Ae. aegypti is unclear (Mackenzie et al. 2004). Gubler (2002) notes 372 that a resurgence of dengue (likely linked to the proliferation of Ae. aegypti) over the past 373 half century is associated with the establishment of peri-urban shantytowns that lack reliable 374 water and sewer services, as well as a large increase in international transportation. 15 375 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Finally, it is noteworthy that increases in population exposure to Ae. aegypti do not 376 necessarily translate into increased risk for dengue or chikungunya virus transmission, due to 377 other factors such as the potential for a vaccine to be developed (Thomas 2015), and other 378 climatic impacts on mosquito ecology (Patz et al. 1998, Liu-Helmersson et al. 2014). Despite 379 these caveats, numerous studies project that 21st century climate change may, overall, 380 facilitate enhanced dengue transmission (Hopp and Foley 2001, Hales et al. 2002, Colon- 381 Gonzalez 2013, Bouzid et al. 2014). 382 We conclude that both climate change and population growth may increase human 383 exposure to the virus vector mosquito Ae. aegypti by the late 21st century. The global 384 population is projected to grow from year 2000 levels of 6053 million (GPW) to 11143 385 million for SSP3 or 8468 million for SSP5 by 2070. The percentage of this total global 386 population exposed to Ae. aegypti for the reference 1950-2000 case is 63%, a value projected 387 to increase to 68-70% by 2061-2080 if population is held constant (GPW; i.e., climate change 388 only), to 77-80% for SSP3 and to 71-74% for SSP5. Developing economies in tropical and 389 subtropical Africa, southern Asia and South America may be particularly impacted by the 390 expansion of year-round Ae. aegypti occurrence patterns, whereas industrialized mid-latitude 391 economies may be impacted by the expansion of seasonal occurrence patterns. The benefits 392 of reduced anthropogenic climate change on human exposure to Ae. aegypti due to taking the 393 RCP4.5 emissions scenario instead of the RCP8.5 scenario may be more pronounced in 394 industrialized countries than in developing and middle income countries. 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 395 Acknowledgements This research was supported by the National Science Foundation 396 (GEO-1010204) and the National Institutes of Health (IR01AI091843). The National Center 397 for Atmospheric Research is sponsored by the National Science Foundation. The SSP data is 398 available on request from Bryan Jones; other data sources are cited within. Lars Eisen of the 399 Centers for Disease Control and Prevention provided invaluable guidance. 400 401 402 Figure Captions Fig. 1 Global extent of Ae. aegypti occurrence patterns (Types 1-4) for (a) the 1950-2000 403 reference period, and changes in occurrence patterns for (b) 2061-2080 RCP4.5 minus the 404 reference and (c) 2061-2080 RCP8.5 minus the reference. Magenta (green) shaded 405 changes are toward more (less) suitable occurrence patterns; e.g., the magenta-shaded "U- 406 >4" category means an unsuitable area becomes Type 4. "N/A" means no change. 407 Fig. 2 Global population affected (millions of persons) for each Ae. aegypti occurrence 408 pattern (Types 1-4) and all patterns combined ("All Types"). Colored bars indicate the 409 ensemble mean value, and the black lines at the top of each bar indicate the range of 410 values among ensemble members. To aid comparison, the bar colors for each Type are 411 identical to the colors used in the global extent map (Fig. 1a). Results are shown for the 412 2061-2080 RCP4.5 and RCP8.5 emissions and concentration scenarios for each of the 413 three population pathways: GPW 2000, SSP3 2070 and SSP5 2070. Violet-colored solid 414 horizontal lines on each panel indicate the population affected for the 1950-2000 415 reference period, assuming the GPW 2000 population; the dashed violet lines are +/- 2 416 standard deviations from the mean. Filled (unfilled) bars indicate populations statistically 417 significantly (insignificantly) different from the 1950-2000 reference period (p<0.05) 17 References Click here to download References monaghan_etal_brace_aedes_paper_references_v5.docx 1 References 2 Anderson GB, Oleson KW, Jones B, Peng RD (2015) Climate change and very dangerous 3 heat waves: projecting frequency of high-mortality heat waves in 82 US communities in 4 2061-2080 under different climate, population and adaptation scenarios. Submitted to 5 Clim Chang 6 Barrera R, Amador M, MacKay AJ (2011) Populaton dynamics of Aedes aegypti and dengue 7 as influence by weather and human behavior in San Juan, Puerto Rico. PLoS Neg Trop 8 Dis 5:e1378. 9 10 11 12 13 Beebe NW, Cooper RD, Mottram P, Sweeney AW (2009) Australia's dengue risk driven by human adaptation to climate change. PLoS Neg Trop Dis 3:e429 Bhatt S, Gething PW, Brady OJ et al (2013) The global distribution and burden of dengue. Nature 496:504-507. doi:10.1038/nature12060 Bouzid M, Colón-González FJ, Lung T, Lake IR, Hunter PR (2014) Climate change and the 14 emergence of vector-borne diseases in Europe: case study of dengue fever. BMC Pub 15 Health 14:781 16 Braks MAH, Honório NA, Lounibos LP, Lourenço-de-Oliveira R, Juliano SA (2004) 17 Interspecific competition between two invasive species of container mosquitoes, Aedes 18 aegypti and Aedes albopictus (Diptera: Culicidae), in Brazil. Ann Entomol Soc Am 19 97:130-139. 20 Campbell LP, Luther C, Peterson AT (2014) Climate change influences on global 21 distributions of dengue and chikungunya virus vectors. Phil Trans R Soc B 370: 22 20140135. doi:10.1098/rstb.2014.0135 23 24 Capinha C, Rocha J, Sousa CA (2014) Macroclimate determines the global range limit of Aedes aegypti. EcoHealth 11:420-428 1 25 Carrington LB, Seifert SN, Willits NH, Lambrechts L, Scott TW (2013) Large diurnal 26 temperature fluctuations negatively influence Aedes aegypti (Diptera: Culicidae) life- 27 history traits. J Med Entomol 50:43-51 28 29 30 Christophers SR (1960) Aedes aegypti (L.), the yellow fever mosquito; its life history, bionomics, and structure. University Press, Cambridge CIESIN (Center for International Earth Science Information Network) (2005) Gridded 31 Population of the World, Version 3 (GPWv3) Data Collection. Columbia University. 32 http://sedac.ciesin.columbia.edu/gpw/index.jsp. Accessed 1 December 2014 33 34 35 36 37 Colón-González FJ, Fezzi C, Lake IR, Hunter PR (2013) The effects of weather and climate change on dengue. PLoS Neg Trop Dis 7:e2503. Ebi KL (2013) Health in the new scenarios for climate change research. Int J Environ Res Pub Health 11:30-46 Eisen L, Monaghan AJ, Lozano-Fuentes S et al (2014) The impact of temperature on the 38 bionomics of Aedes (Stegomyia) aegypti, with special reference to the cool geographic 39 range margins. J Med Entomol 51:496-516 40 Eisen L, Moore CG (2013) Aedes (Stegomyia) aegypti in the continental United States: A 41 vector at the cool margin of its geographic range. J Med Entomol 50:467-478 42 Ernst KC, Haenchen S, Dickinson K et al (2015) Awareness and support of release of 43 genetically  modified  “sterile”  mosquitoes,  Key  West,  Florida,  USA.  Emerg  Infect  Dis in 44 press. doi:10.3201/eid2102.141035 45 Focks DA, Haile DG, Daniels E, Mount GA (1993) Dynamic life table model for Aedes 46 aegypti (Diptera: Culicidae): analysis of the literature and model development. J Med 47 Entomol 30:1003-1017 2 48 Gubler DJ (1998) Dengue and dengue hemorrhagic fever. Clin Microbiol Rev 11:480–496 49 Gubler DJ (2002) The global emergence/resurgence of arboviral diseases as public health 50 51 problems. Arch Med Res 33:330-342 Hales S, De Wet N, Maindonald J, Woodward A (2002) Potential effect of population and 52 climate changes on global distribution of dengue fever: an empirical model. Lancet 53 360:830-834 54 55 56 57 Halstead SB (2015) Reappearance of chikungunya, formerly called dengue, in the Americas. Emerg Infect Dis 21:557-561 Held IM, Soden BJ (2006) Robust responses of the hydrological cycle to global warming. J Clim 19:5686-5699 58 Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution 59 interpolated climate surfaces for global land areas. Int J Climatol 25:1965-1978 60 61 62 63 64 65 66 67 Hopp MJ, Foley JA (2001) Global-scale relationships between climate and the dengue fever vector, Aedes aegypti. Clim Chang 48:441-463 Hurrell JW, Holland MM, Gent PR et al (2013) The Community Earth System Model: a framework for collaborative research. Bull Am Meteorol Soc 94:1339-1360 Jiang  L,  O’Neill  BC  (2015)  Global  urbanization  projections  for  the  Shared  Socioeconomic Pathways. Glob Env Chang. doi:10.1016/j.gloenvcha.2015.03.008 Jones  B,  O’Neill  BC  (2015)  Paper  on  decadal global gridded population projections for the 21st century based on Shared Socioeconomic Pathways. In preparation. 68 Jones B, O’Neill BC, Oleson K (2015) Avoiding population exposure to heat-related 69 extremes: Demographic change vs climate change. Submitted to Clim Chang 3 70 Kay JE, Deser C, Phillips A et al (2014) The Community Earth System Model (CESM) Large 71 Ensemble Project: A community resource for studying climate change in the presence of 72 internal climate variability. Bull Am Meteorol Soc. doi://10.1175/BAMS-D-13-00255.1 73 KC S, Lutz W (2014) The human core of the shared socioeconomic pathways: Population 74 scenarios by age, sex and level of education for all countries to 2100. Glob Env Chang. 75 doi:10.1016/j.gloenvcha.2014.06.004 76 Kearney M, Porter WP, Williams C, Ritchie S, Hoffmann AA (2009) Integrating biophysical 77 models and evolutionary theory to predict climatic impacts on species' ranges: the 78 dengue mosquito Aedes aegypti in Australia. Funct Ecol 23:528-538 79 Khormi HM, Kumar L (2014) Climate change and the potential global distribution of Aedes 80 aegypti: spatial modelling using geographical information system and CLIMEX. Geosp 81 Health 8:405-415 82 Koenraadt CJM, Harrington LC (2008) Flushing effect of rain on container-inhabiting 83 mosquitoes Aedes aegypti and Culex pipiens (Diptera: Culicidae). J Med Entomol 84 45:28-35 85 86 87 Kroeger A, Nathan MB (2006) Dengue: setting the global research agenda. Lancet 368:2193-2195 Liu-Helmersson J, Stenlund H, Wilder-Smith A, Rocklöv J (2014) Vectorial capacity of 88 Aedes aegypti: effects of temperature and implications for global dengue epidemic 89 potential. PloS ONE 9:e89783 90 91 92 93 Lounibos LP, Suarez S, Menendez Z et al (2002) Does temperature affect the outcome of larval competition between Aedes aegypti and Aedes albopictus. J Vector Ecol 27:86-95 Lucio PS, Degallier N, Servain J et al (2013) A case study of the influence of local weather on Aedes aegypti (L.) aging and mortality. J Vector Ecol 38:20-37 4 94 Mackenzie JS, Gubler DJ, Petersen LR (2004) Emerging flaviviruses: the spread and 95 resurgence of Japanese encephalitis, West Nile and dengue viruses. Nature medicine 96 10:S98-S109 97 Marsha A, Sain SR, Heaton MJ, Monaghan AJ, Wilhelmi OV (2015) Climate change 98 influences on extreme heat mortality in Houston, Texas. Submitted to Clim Chang 99 Moji M, Khamboonruang C, Choochote W, Suwanpanit P (1988) Ovitrap surveys of dengue 100 vector mosquitoes in Chiang Mai, northern Thailand: seasonal shifts in relative 101 abundance of Aedes albopictus and Ae. aegypti. Med Vet Entomol 2:319-324 102 103 Morin CW, Comrie AC, Ernst KC (2013) Climate and dengue transmission: evidence and implications. Env Health Perspect 121:1264-1272 104 Morin CW, Monaghan AJ, Hayden MH, Barrera R, Ernst K (2015) Meteorologically driven 105 simulations of dengue epidemics in San Juan, PR PLoS Neg Trop Dis 9:e0004002 106 Morrison TE (2014) Reemergence of chikungunya virus. J Virol 88:11644-11647 107 Nasci RS (2014) Movement of chikungunya virus into the Western Hemisphere. Emerg 108 109 Infect Dis 20:1394 O'Neill BC, Gettleman A (Eds.) (2015) The Benefits of Reduced Anthropogenic Climate 110 changE (BRACE): Introduction to the special issue on the BRACE Project. Submitted to 111 Clim Chang 112 O’Neill  BC,  Ebi KL, Kemp-Benedict E et al (2015) The roads ahead: narratives for the 113 Shared Socioeconomic Pathways. Glob Environ Chang in press. 114 doi:10.1016/j.gloenvcha.2015.01.004 5 115 Parham PE, Waldock J, Christophides GK et al (2015) Climate, environmental and socio- 116 economic change: weighing up the balance in vector-borne disease transmission. Phil 117 Trans R Soc B 370: 20130551 118 Patz JA, Martens WJ, Focks DA, Jetten TH (1998) Dengue fever epidemic potential as 119 projected by general circulation models of global climate change. Env Health Perspect 120 106:147 121 122 Pialoux G, Gauzere B-A, Jaureguiberry S, Strobel M (2007) Chikungunya, an epidemic arbovirosis. Lancet Infect Dis 7:319-327 123 Rogers DJ (2015) Dengue: recent past and future threats. Phil Trans R Soc B 370:20130562. 124 Sanderson B, Tebaldi C, Knutti R (2015) On the dependency of climate variability and 125 126 127 128 129 130 131 132 133 134 extremes with mean climate state. Submitted to Clim Chang Tabachnick WJ Powell JR (1979) A world-wide survey of genetic variation in the yellow fever mosquito, Aedes aegypti. Genet Res 34:215-229 Taylor KE, Stouffer RJ, Meehl GA (2012) An Overview of CMIP5 and the Experiment Design. Bull Am Meteorol Soc 93:485–498 Thomas SJ (2015) Preventing Dengue—Is the Possibility Now a Reality? New Engl J Med. Doi:10.1056/NEJMe1413146 van Vuuren DP, Edmonds J, Kainuma M et al (2011) The representative concentration pathways: an overview. Clim Chang 109:5-31 Williams CR, Mincham G, Ritchie SA, Viennet E, Harley D (2014) Bionomic response of 135 Aedes aegypti to two future climate change scenarios in far north Queensland, Australia: 136 implications for dengue outbreaks. Parasites and Vectors 7:447 6 137 138 139 140 WHO (World Health Organization) (2009) Dengue: guidelines for diagnosis, treatment, prevention and control – New edition. WHO Press, Geneva Xu Y, Lamarque J-F, Sanderson B (2015) The importance of aerosol scenarios in projections of future heat extremes. Submitted to Clim Chang 141 7 Figures a) 1950-2000 (Reference) Unsuitable Type 1 Type 2 Type 3 Type 4 b) 2061-2080 (RCP4.5 minus Reference) U->4 4->3 3->2 2->1 N/A 1->2 2->3 3->4 4->U c) 2061-2080 (RCP8.5 minus Reference) U->4 4->3 3->2 2->1 N/A 1->2 2->3 3->4 4->U Fig. 1 Global Population Affected (millions) RCP4.5 RCP8.5 RCP4.5 RCP8.5 RCP4.5 RCP8.5 GPW SSP3 SSP5 Scenario, 2061 -2080 Fig. 2 Tables 1 Table 1. Thresholds used to define the occurrence patterns for Ae. aegypti. Type 1 Type 2 Type3 Type 4 Variable Metric ≥ ≤ ≥ ≤ ≥ ≤ ≥ ≤ Monthly Air Annual Mean 25.4 - 23.7 - 16.8 - 14.5 - Temperature Annual Min 24.6 - 14.2 - 4.2 - 0.8 - (oC) Annual Max - - - - - - - - Annual Range - 4.1 - 19.3 - 23.9 Monthly Annual Mean 1498 - 705 - 675 - 186 - Precipitation Annual Min - - - - - - - - (mm) Annual Max 304 - 148 - 134 - 27 - Annual Range 145 - 96 - 70 - 17 - 26.7 2 3 1 4 Table 2. Change in land surface area (millions km2) for each type of Ae. aegypti occurrence pattern, for 2061-2080 versus 1950-2000, for each 5 region and RCP scenario. Pixels were weighted by the sin of the latitude when calculating surface area.* Region Africa Asia Australia Europe N. America S. America Global 6 1950-2000 Reference 0.2 1.1 0.0 0.0 0.1 3.5 4.9 Type1 Δ  2061-2080 RCP4.5 RCP8.5 0.5 0.9 0.7 0.9 0.1 0.1 0.0 0.0 0.1 0.1 1.4 0.2 2.7 2.2 Type 2 1950-2000 Δ  2061-2080 Baseline RCP4.5 RCP8.5 7.3 1.9 3.2 4.4 0.3 0.6 0.9 0.2 0.3 0.0 0.0 0.0 0.8 0.2 0.3 5.6 0.2 1.9 19.0 2.9 6.3 Type 3 1950-2000 Δ  2061-2080 Baseline RCP4.5 RCP8.5 5.1 -1.9 -3.4 2.9 -0.2 -0.5 0.3 0.0 -0.1 0.0 0.0 0.1 1.1 0.3 0.3 3.2 -0.9 -1.3 12.5 -2.7 -4.8 Type 4 All Types 1950-2000 Δ  2061-2080 1950-2000 Δ  2061-2080 Reference RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5 Africa 7.0 -0.1 -0.2 19.5 0.3 0.4 Asia 3.2 0.4 1.5 11.6 1.3 2.6 Australia 4.9 0.7 0.7 6.1 1.0 1.0 Europe 0.4 0.5 1.1 0.4 0.6 1.2 N. America 2.5 0.4 0.8 4.5 1.0 1.5 S. America 2.6 -0.4 -0.3 15.0 0.4 0.5 Global 20.5 1.6 3.6 56.9 4.5 7.3 *underlined italicized values indicate statistically significant (p<0.05) changes for 2061-2080 compared to 1950-2000. 2 7 Table 3. Change in population for "All Types" of Ae. aegypti occurrence patterns for 2061-2080 versus 1950-2000, for each region, SSP 8 pathway, and RCP scenario, as well as the difference of the RCP8.5 and RCP4.5 scenarios.* 9 10 11 Reference RCP4.5 RCP8.5 RCP8.5-minus-RCP4.5 1950-2000 Δ  2061-2080 Δ  2061-2080 Δ  2061-2080 Region GPW GPW SSP3 SSP5 GPW SSP3 SSP5 GPW SSP3 SSP5 Africa 664 16 2093 1023 19 2104 1029 3 10 6 Asia 2538 145 2172 747 202 2334 846 58 162 99 Australia 17 4 17 43 5 18 45 1 1 2 Europe 44 64 48 113 129 106 194 65 58 81 N. America 253 56 245 236 84 280 290 28 35 54 S. America 277 13 230 70 20 243 78 7 13 9 Global 3794 298 4805 2232 460 5084 2483 162 279 251 *underlined italicized values indicate statistically significant (p<0.05) changes for 2061-2080 compared to 1950-2000 for the RCP4.5 and RCP8.5 scenarios. For the RCP8.5-minus-RCP4.5 scenario, italicized underlined values indicate statistically significant differences (p<0.05) for the RCP8.5 versus RCP4.5 scenario 3 Supplementary Material 1 Click here to access/download Supplementary Material monaghan_etal_brace_aedes_paper_ESM1_v5.docx Supplementary Material 2 Click here to access/download Supplementary Material monaghan_etal_brace_aedes_paper_ESM2_v5.docx