I have explained that we can do much better than one-size-fits-all thinking as we work to make buildings more safe from climate change, and my previous post introduced a first concept to get past this limitation, which is to move beyond averages. This post dives deeper, into non-linear relationships and critical thresholds. These two effects work together to mean that a small change in average temperature decade-over-decade translates into a much greater impact on climate-related fatalities.
In addition to swamping the existence of potentially catastrophic events, the small average temperature change predicted by climate models can be misleading in a more subtle way: the relationships between temperature change and the downstream effects of this are nonlinear. This means that even a small change in temperature can result in consequences that get disproportionately bigger as the temperature increases.
The relationship between increasing temperature and the consequent death rate among the elderly follows this pattern: Each incremental increase in air temperature by 1°F above 95°F raises the death rate by about 0.3 deaths per hundred thousand people aged 65 and older. So, in the San Francisco Bay Area, for example, where about 8 million people live, 14.2% of whom are above 65 years of age, raising the temperature by one degree for one day will likely result in approximately 4 additional deaths. Raise it by 2 degrees and the death toll increases to 8. Leave it at that temperature for two days instead of one, and now 16 people will die. The higher the temperature climbs above 95, the more people die every day. The more days above 95 during any given period, the more people will die.
The San Francisco Bay Area: a case in point
The Bay Area near San Francisco, California has historically had a very mild, temperate climate. Winters are not bitterly cold, nor are summers blazingly hot. In fact, during the entire decade from 2000-2010, San Francisco experienced only 1 day above 100°F, and only 13 days above 95°F. There have been very few days where the temperature poses a mortal risk to anyone.
Raising the average temperature by only one degree Fahrenheit, as occurred in the Bay Area during the following decade, increased the number of days above 95°F by only 2—totaling 15 between 2010-2020. However, the severity and distribution of these hot days changed. Between 2000-2010, the majority of hot days were either 95°F or 96°F. The following decade, they were more evenly spread from 95°F to 104°F.
Since the mortality rate goes up as the temperature goes up, although the number of hot days didn’t change very much from 2000-2020, the fact that the hottest days were hotter during the second decade disproportionately increased the death rate. This jumped from 16 to 27 per 100,000, despite an average temperature rise of only 1.4°F over the decade, and only two more hot days.
The table and chart below show these Bay Area temperatures as measured for the decades ending in 2010 and 2020, along with the predicted results based on model data for decades ending between 2030 and 2050. The non-linear relationships between average temperature rise, and both the number of hot days per year, and the mortality rate, are clear. Despite the average temperature rising by less than 6°F over the 50-year period, the elderly death rate due to heat stress has increased to more than eight times its original value.
Examining the data makes it clear how and why this result comes about. The histogram below shows that each successive decade has more hot days, and that those days are hotter, than the previous decades. Together, these two effects multiply to produce a death rate that increases more steeply with each decade.
There is an additional reason why the apparently small change in average temperature that climate models predict is deceptive. When a system is near a “critical value”—that is, a value which acts as a dividing line between one kind of behavior and another—a very small change is all that’s needed to push it over that line, after which the system may behave in dramatically different ways.
In our case, the temperature of 95°F is such a value. Stay below it, and very few people die. Rise above it and people start dying, and the higher the temperature rises, the greater the mortality. The chart below shows why a 5.7°F rise in average temperature over 50 years can lead to such deadly results. Prior to 2010, there were almost no days that crossed the critical threshold. But nudge the average temperature up even a small amount, and the relative fraction of the hottest days that are now over the critical threshold increases many-fold. Deaths that were once rare are now commonplace
How can we make buildings safer, so that critical thresholds and nonlinear effects have a smaller impact on their occupants? Please subscribe for future posts about how building information management, AI, and smart decisions can stop climate change and biohazards from turning buildings into death traps.