I guess sometimes you just have to throw the officially acknowledged lack of regional skill for GCM’s out the window~ctm
A major challenge in articulating human dimensions of climate change lies in translating global climate forecasts into impact assessments that are intuitive to the public. Climate-analog mapping involves matching the expected future climate at a location (e.g., a person’s city of residence) with current climate of another, potentially familiar, location – thereby providing a more relatable, place-based assessment of climate change. For 540 North American urban areas, we used climate-analog mapping to identify the location that has a contemporary climate most similar to each urban area’s expected 2080’s climate. We show that climate of most urban areas will shift considerably and become either more akin to contemporary climates hundreds of kilometers away and mainly to the south or will have no modern equivalent. Combined with an interactive web application, we provide an intuitive means of raising public awareness of the implications of climate change for 250 million urban residents.
Within the lifetime of children living today, the climate of many regions is projected to change from the familiar to conditions unlike those experienced in the same place by their parents, grandparents, or perhaps any generation in millenia1,2. While scientists share great concern for the expected severe impacts of climate change, the same is not necessarily true of the general public3,4,5. At the same time, decision makers have not formalized climate adaptation plans for a large proportion of major cities6, and existing efforts often are considered insufficient to avoid social, environmental, and economic consequences of climate change7.
Disconnects between the potential threats of climate change and societal action arise from multiple factors4,5,8, but changing how people perceive and conceptualize climate change is considered key to improving public engagement4,5,8. For example, it is difficult for people to identify with the abstract, remote, descriptive predictions of future climate used by scientists (e.g., a 3 °C increase in mean global temperature). Translating and communicating these abstract predictions in terms of present-day, local, and concrete personal experiences may help overcome some barriers to public recognition of the risks (and opportunities) of climate change9,10. Given that most humans reside in urban areas and urban populations are considered highly sensitive to climate change11, it is important to assess what climate change could mean for urban areas and to communicate the magnitude and uncertainty of these expected changes in intuitive ways.
Climate-analog mapping is a statistical technique that quantifies the similarity of a location’s climate relative to the climate of another place and/or time12,13,14,15. When considered in the context of assessing and communicating exposure to future climate change, climate-analog mapping can be viewed as a form of forecasting by analogy16,17 that translates the descriptive and abstract (i.e., scientific forecasts of future climate) into the familiar (i.e., present-day climate of a known location). Veloz et al.18 used climate-analog mapping to find contemporary climatic analogs for projected future climates for the U.S. state of Wisconsin, while Rohat et al.19 used similar methods to quantify and communicate the implications of climate change for 90 European cities. Climate-analog mapping is gaining popularity as a means to communicate climate change impacts20,21, and more robust methods for measuring climatic similarity between places and times have been recently developed22.
Here we use climate-analog mapping and an interactive web application (available at https://tinyurl.com/urbanclimate) to characterize and communicate how climate change may impact the lives of a large portion of the populations of the United States and Canada. Collectively, the 540 urban areas we analyze in this study include approximately 250 million inhabitants, including >75% of the population of the United States and >50% of the population of Canada. For each urban area, we mapped the similarity between that city’s future climate expected by the 2080s (mean of the period 2070–2099)23 and contemporary climate (representative of mean conditions for 1960–1990)24 in the western hemisphere north of the equator (Supplementary Figure 1). We identified climatic analogs using sigma dissimilarity22, a statistical measure that accounts for correlations between climate variables, incorporates historical interannual climatic variability (ICV), and converts multidimensional climatic distances to percentiles of a probability distribution of these distances. A sigma dissimilarity equal to 0 (i.e., 0σ) would indicate identical climates, or a perfect analog. We considered values of ≤2σ between an urban area’s future climate and its most similar contemporary climate to be a representative analog. Values >4σ represent extreme differences between future climate and contemporary climate within the study domain, which we interpret as novel future climatic conditions22 and a poor analog. In this sense, sigma dissimilarity serves as both an indicator of climate novelty and a measure of the strength of analogy between an urban area’s future climate and its best contemporary climate match.
We calculated sigma dissimilarity using minimum and maximum temperature and total precipitation for the four climatological seasons (12 climate variables total). For 2080’s climate, we selected two emission trajectories or Representative Concentration Pathways (RCPs)25, unmitigated emissions (RCP8.5) and a mitigation scenario (RCP4.5)26, and 27 different earth system models (ESMs), for a total of 2 RCPs × 27 ESMs = 54 future climate scenarios (Supplementary Table 1). Here we emphasize results for the ensemble means of 2080’s climate calculated by averaging across the 27 climate projections for each RCP.
For each future climate scenario, we calculated sigma dissimilarity between each urban area’s future climate and every contemporary climate pixel in the study domain. We mapped the resulting sigma values to create a climate similarity surface and identified the pixel with the minimum sigma dissimilarity. This pixel represents the best contemporary climatic analog to 2080’s climate for that urban area and climate scenario, again noting that values >2σ increasingly characterize novel climates rather than representative analogs.
We find that if emissions continue to rise throughout the 21st century, climate of North American urban areas will become, on average, most like the contemporary climate of locations 850 km away and mainly to the south, with the distance, direction, and degree of similarity to the best analog varying by region and assumptions regarding future climate. For many urban areas, we found substantial differences between future climate and the best contemporary climatic analog, underscoring that by the 2080s many cities could experience novel climates with no modern equivalent in the study domain. In addition to the summaries we report here, we visualize climate analogs for all 540 urban areas and 54 future climate scenarios using an interactive web-based application (available at https://tinyurl.com/urbanclimate) that provides a means to communicate abstract forecasts of future climate in terms that are more locally relevant to the nearly 250 million people who call these urban areas home.
HT/ Clyde Spencer