This is the last post is in the Niskanen Center’s response to Benjamin Zycher’s case against climate action, which he forwarded in advance of the COP-21 meeting in Paris. While that meeting is well past, the arguments that Zycher advances are worth our attention given their omnipresence in the skeptic community.
Today, we’ll consider how climate change might contribute to weather phenomena. In his article, Zycher claims the following:
- There is no apparent trend in tornado count or intensity.
- Data show that tropical cyclones are not worsening.
- Wildfires in the United States are not increasing.
- The U.S. drought index shows no trend.
- U.S. flooding is not correlated with increasing GHG concentrations.
Each of these phenomena constitute significant risks to life, infrastructure, and property in the United States and around the world. As we consider the risks created by climate change, it is important to ask if a warming world has been increasing the severity of these phenomena—or if it will in the future.
Zycher finds limited support for the idea that climate change is making any of these worse. Let’s have a look at the evidence.
Zycher uses historical tornado data to rebut the “the common assertion that increasing GHG concentrations are exacerbating U.S. tornado activity.” Fair enough on the data, but what study or assessment embarked on such an argument?
The major assessments are equivocal on the question of whether CO2-driven greenhouse warming has already affected tornado activity and whether the total affect of warming will significantly alter tornado activity. In its special report on extreme events and disasters, the IPCC made two important points:
- There is low confidence in observed trends in small spatial-scale phenomena such as tornadoes and hail because of data inhomogeneities and inadequacies in monitoring systems.
- There is low confidence in projections of small spatial-scale phenomena such as tornadoes and hail because competing physical processes may affect future trends and because current climate models do not simulate such phenomena.
Low confidence is IPCC speak for we can’t really say, because we have limited observational knowledge of tornado counts and intensity in the past, long-term trends cannot be identified with confidence.
Zycher invokes two measures of tropical cyclone activity—frequency of tropical cyclones and accumulated cyclone energy (a measure of intensity)—to argue that “data do not support the argument that tropical storms are worsening because of increasing GHG concentrations.”
As with tornadoes, the challenges in long-term record keeping limit knowledge about changing frequency or intensity of tropical cyclones since humans started affecting the climate. In IPCC assessments, conclusions about tropical cyclone trends are also made with low confidence.
But that does not prove that climate change has had no effect on tropical cyclones. Within the satellite era, there have been some changes that might be expected with climate change. The latitude at which tropical cyclones reach their peak intensity has moved poleward, consistent with models of global warming moving the boundaries of the tropics poleward.
Scientists predict increased wildfire risk in North America under global warming, as some areas are expected to get both hotter and drier.
Zycher presents data on the number of fires to say that these aren’t changing, but he is looking in the wrong place. When we look at the area burned (c.f. Fig 1) a recent trend is apparent. The number of large fires in the Western United States is increasing; and, globally, the mean fire season length has increased by 20 percent from 1979 to 2013.
Figure 1: Area of wildfires in the US, 1960-2014. (Smoothed curve in red with 95% range indicated).
Zycher shows the average Palmer Drought Severity Index (PDSI) for the contiguous United States and says that it is unchanging. But it is not clear that this metric matters, because regional details are important. The general rule under global warming is that wet areas get wetter, while dry areas get drier. So how would average PDSI tell one anything about that?
Globally, drought conditions (defined as the land area meeting a drought threshold, not an index average) are increasing and expected to increase in the future, but attributing trends in drought to climate change can only be done with low confidence. But scientists are increasingly looking at individual events to investigate climate change as a driver of drought events. For instance, one recent study showed that climate change contributed to the severity of the ongoing drought in California, increasing its severity by 8-27 percent in 2012-2014.
Zycher cites analysis of stream gauge data to show that flooding in the United States is not correlated with atmospheric CO2 concentrations. Zycher’s conclusion is not too far off from what the IPCC reported in its last assessment: “Few discernible trends in flooding have been observed in the USA. Changes in the magnitude or frequency of flood events have not been attributed to climate change.”
While extreme precipitation events appear to be increasing, that has not apparently translated into stronger flooding events at stream gauges. However, more recent analysis shows that floods are more frequent, while not necessarily higher, in the central United States.
Zycher points out that the only statistical relationship of any merit is a reduction in flooding in the Southwest and Rocky Mountain regions. The authors of the cited paper offer an explanation:
In particular, in the southwestern region, the warming that has taken place is likely to be causing decreased winter snowpacks in some of the watersheds and this decrease may be contributing to a decreased potential for flooding.
Such a reduction in flooding might be a fortunate side effect of decreasing snowpack, but the impact of climate should not be dismissed. Especially after Zycher had just made an effort to dismiss a relationship between climate changes and drought.
Sea level rise illustrates an important point about Zycher’s post and the wide array of evidence that he cites as “the actual evidence on the relationship between increasing atmospheric greenhouse gas (GHG) concentrations and such climate phenomena as storms, polar ice, and droughts.”
Left out of Zycher’s post is any examination of the underlying physical mechanisms that help us understand how each of these phenomena change in a warming world. An understanding of risk need not come only from correlation. Physical understanding can help.
As an example, take coastal flooding. It is made worse by sea level rise. In part 2 of this series, we looked at the human contribution to sea level rise and found it to be substantial. A recent report on U.S. coastal flooding events (at 27 tide gauge sites with good data) found that 67 percent of coastal flooding days since 1950 were driven by climate change. As sea levels continue to rise over the next 15-30 years, coastal flooding will most likely inundate cities around the coastal United States regularly at high tides.
Some risks are more predictable (or certain):
- Increased coastal flooding is practically assured.
- Global warming will affect moisture in the atmosphere, likely leading to drought in some areas and increased precipitation in others.
- Areas of increased drought and warming will experience increased risk of wildfire.
Others are still ambiguous:
- Climate modeling indicates that a warmer and moister atmosphere will create conditions that favor convective storms and tornadoes, but this is far from certain.
- Theoretical and modeling exercises show that tropical cyclones have the potential to increase in frequency and/or intensity. This has been modeled for both low and high-end climate change scenarios and is supported theoretically, but the details are still unclear. We might not detect such changes until mid-century or later. In terms of risk to life and property, wea level rise will likely increase the damages associated with hurricanes, even if the storms themselves do not change. When changing tropical cyclone activity is input into private-sector flood risk models, it increases expected losses by about a factor of three for the next century.
No weather event is unaffected by climate change. But the effects of climate changes on storms and drought and other phenomena varies by region. Given the episodic nature of weather, changes that are predicted by physics may not be statistically detectable for some time. Even with physical and theoretical understanding, predictions for some phenomena have substantial uncertainty.
Zycher’s analysis shows little appreciation of how climate change is presently affecting weather phenomena or how it will in the future.
We have spent a few blog posts considering the claims Zycher made before the Paris climate conference took place and before the agreement was settled. His analysis on the connection between CO2 emissions and temperature changes, ice melt, sea level rise, and weather phenomena was incomplete and often poorly informed. There is much more evidence for the existence, risks, and human causes of climate change than Zycher considered.
We conclude that there is little scientific doubt that humans are behind much of the climate change that we have observed. What the future temperature (or weather phenomena) will hold under continued human influence is deeply uncertain. But these changes present real risks to life and property.
But we do agree on some matters. Zycher highlighted in his post that the emissions reductions pledged at Paris would do little to reduce temperature change at the end of this century. That is likely so, even if all countries are able to meet their pledges. In the United States, how the Administration plans to meet its pledge is ambiguous. How other countries will respond to that ambiguity is unknown. A carbon price regime, hopefully via carbon taxes, would offer simplicity and perhaps much more ambition.