Public Perception of Climate Change and the New Climate Dice
James Hansen, Makiko Sato, Reto Ruedy

Summary. Should the public be able to recognize that climate is changing, despite the notorious variability of weather and climate from day to day and year to year? We investigate how the probability of unusually warm seasons has changed in recent decades, with emphasis on summer, when changes are likely to have the greatest practical effects. We show that the odds of an unusually warm season have increased greatly over the past three decades, but also the shape of the frequency distribution has changed so as to enhance the likelihood of extreme events. A new category of hot summertime outliers, more than three standard deviations (3σ) warmer than climatology, has emerged, with the occurrence of these outliers having increased 1-2 orders of magnitude in the past three decades. Thus we can state with a high degree of confidence that extreme summers, such as those in Texas and Oklahoma in 2011 and Moscow in 2010, are a consequence of global warming, because global warming has dramatically increased their likelihood of occurrence.

We illustrate observed variability of seasonal mean surface air temperature anomalies in units of standard deviations, including comparison with the normal distribution (“bell curve”) that the lay public may appreciate. We take 1951-1980 as an appropriate base period, because temperatures then were within the Holocene range to which humanity and other planetary life are adapted. In contrast, we infer that global temperature is now above the Holocene range, as evidenced by the fact that the ice sheets in both hemispheres are shedding mass (1) and sea level is rising at a rate [more than 3 mm/year or 3 m/millennium (2)] that is much higher than the rate of sea level change during the past several millennia.

The frequency of occurrence of local summer-mean temperature anomalies was close to the normal distribution in the 1950s, 1960s and 1970s in both hemispheres (Fig. P1A, B). However, in each subsequent decade the distribution shifted toward more positive anomalies, with the positive tail (hot outliers) of the distribution shifting more than the negative tail. The temporal change of the anomaly distribution for the contiguous United States (Fig. P1C) is similar to the global change, but much noisier because the contiguous U.S. covers only ~1.5% of the globe.

Winter warming exceeds that in summer, but the standard deviation of seasonal mean temperature at middle and high latitudes is much larger in winter (typically 2-4°C) than in summer (typically ~1°C). Thus the shift of the anomaly distribution, in the unit of standard deviations, is less in winter than in summer (Fig. P1D).

A concept of “climate dice” was suggested (3) to describe the stochastic variability of local seasonal mean temperature, with the implication that the public should recognize the existence of global warming once the dice become sufficiently “loaded” (biased). Specifically, the 10 warmest summers (Jun-Jul-Aug in the Northern Hemisphere) in the 30-year period ofclimatology (1951-1980) define the “hot” category, the 10 coolest the “cold” category, and the middle 10 the “average” summer. Thus it was imagined that two sides of a six-sided die were colored red, blue and white for these respective categories. The divisions between “hot” and “average” and between “average” and “cold” occur at +0.43σ and -0.43σ for a normal distribution of variability.

Temperatures simulated in a global climate model reached a level such that four of the six sides of the climate dice were red in the first decade of the 21st century for greenhouse gas scenario B (3), which is an accurate approximation of actual greenhouse gas growth [(4), updates at http://www.columbia.edu/~mhs119/GHG_Forcing/]. We find that actual summer-mean temperature anomalies over global land during the past decade averaged about 75% in the “hot category”, thus midway between four and five sides of the die were red, which is reasonably consistent with expectations.

A more important change is the emergence of a subset of the hot category, extremely hot outliers, defined as anomalies exceeding +3σ. The frequency of these extreme anomalies is about 0.13% in the normal distribution, and thus a typical summer in the period of climatology would have only about 0.1-0.2% of the globe covered by such hot extremes. We show that during the past several years the portion of global land area covered by summer temperature anomalies exceeding +3σ has averaged about 10%, thus an increase by about a factor of 50 compared to the period of climatology. Recent examples of summer temperature anomalies exceeding +3σ include the heat wave and drought in Oklahoma, Texas and Mexico in 2011 and a larger region encompassing much of the Middle East, Western Asia and Eastern Europe, including Moscow, in 2010.

The question of whether these extreme hot anomalies are a consequence of global warming is commonly answered in the negative, with an alternative interpretation based on meteorological patterns. For example, an unusual atmospheric “blocking” situation resulted in a long-lived high pressure anomaly in the Moscow region in 2010, and a strong La Nina in 2011 may have contributed to the heat and drought situation in the southern United States and Mexico. However, such meteorological patterns are not new and thus as an “explanation” fail to account for the huge increase in the area covered by extreme positive temperature anomalies. Specific meteorological patterns help explain where the high pressure regions that favor high temperature and drought conditions occur in a given summer, but the unusually great temperature extremities and the large area covered by these hot anomalies is a consequence of global warming.

This attribution is important, because we can project with a high degree of confidence that the area covered by extremely hot anomalies will continue to increase during the next few decades and even greater extremes will occur. The decade-by-decade shift to the right of the temperature anomaly frequency distribution (Fig. P1) will continue, because Earth is out of energy balance, more solar energy absorbed than heat radiation emitted to space (5), and it is this imbalance that drives the planet to higher temperatures. Even an extremely optimistic scenario, with fossil fuel emission reductions of 6%/year beginning in 2013, results in global temperature rising to almost 1.2°C relative to 1880-1920, which compares to a current level ~0.8°C.

We argue that it is important to keep the base period defining climatology fixed. Shifting the base period continually to the most recent three decades hides the increasing variability that we found. A base period prior to 1980 avoids this problem and yields a climatology within the global temperature range of the Holocene, to which nature and human civilization are adapted.

Practical effects of the increasingly loaded climate dice are likely to occur via amplification of extremes of the water cycle. Higher temperatures exacerbate hot dry conditions, but higher temperatures also increase the amount of water vapor that the atmosphere can hold. Increased water vapor leads to heavier rainfall and floods as well as the potential for stronger storms driven by latent heat including thunderstorms, tornadoes and tropical storms. We cite data suggesting that such climate impacts are already underway, but because of the small spatial scale of many of these phenomena it is necessary to gather more extensive homogeneous hydrologic data to assess ongoing global change. Such assessment is important because of potential effects on humans and other species, as it has been estimated that continued business-as-usual fossil fuel emissions and global warming could result by the end of the century in 21-52% of the species on Earth being committed to extinction IPCC (6).