Preliminary Assessment of Climate Factors Contributing to the Extreme 2011 Tornadoes

Draft Assessment - Updated: 08-July-2011

Disclaimer: This draft is an evolving research assessment and not a final report. Comments are welcome. For further information, contact Dr. Martin Hoerling (
Overview | El Niño/La Niña | Climate Change | Q & A | References

Climate Change and Tornadoes

There is a heightened sense of interest, concern, and urgency to understand any potential role climate change may be having on extreme events such as the April 2011 tornado outbreaks.

One question on many minds concerns the role of anthropogenic climate change. Two recent national and international assessment reports have summarized the existing state of knowledge on climate change and tornadoes. According to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) :

    "There is insufficient evidence to determine whether trends exist in.....small-scale phenomena such as tornadoes, hail, lightning and dust-storms." 2

The US Climate Change Synthesis Report SAP 3.3 concludes that:

    "The data used to examine changes in the frequency and severity of tornadoes and severe thunderstorms are inadequate to make definitive statements about actual changes." 3

The report also concludes that:

    "There were no significant changes in the high-intensity end of these distributions from the 1950s through the 1990s, although the distribution from 2000 and later may differ."
a. Illinois Tornado Chart
b. Illinois Tornado F0 Chart
Fig. 1. The annual number of tornadoes per year in Illinois since 1950, regardless of strength (a) and F0 only (b). (Source: State Climatologist Office for Illinois)

The difficulties in assessing change directly from the history of tornado data is emphasized in both reports. A simple illustration of the problem is provided by the 1950-2005 time series of the number of tornadoes in Illinois, as constructed by the Illinois State Climatologist Office 4; the number of tornado counts has almost doubled since 1990. (Fig. 1a) The post-1990 increase in counts is almost entirely associated with the counts in weak (F-0) tornadoes (Fig. 1b). Careful appraisal of historical tornado outbreaks has led scientists to recognize that trends in tornado occurrences before 1970 are likely more influenced by non-meteorological factors6. Even after 1970, non-meteorological factors may have contributed to more reporting of tornadoes, including changed instrumentation (e.g., the implementation of the WSR-88D Doppler radar after 1990), shifts in population, and public awareness via enhanced spotter networks that may have contributed to more reporting.5 Indeed, in 2011, the high loss of life was partly related to the unfortunate circumstance that the severe tornados happened to strike large urban population centers. At this time, the historical tornado data are not of a quality to permit a rigorous appraisal of trends, although the database of stronger tornadoes (F2+) is more reliable, especially over the last 30 years.

Can variations in tornadoes be inferred from variables other than the tornado counts themselves? A recent special report by the WMO Working Group and Expert Team on Detection and Attribution Related to Anthropogenic Climate Change7 offers guidelines on attribution research. These guidelines propose that a first step toward a rigorous assessment is to determine whether a change in an event, or in variables closely related to such events, has been detected. Thus, whereas the tornado data themselves is not reliable enough for rigorous assessment of physical trends, perhaps trends in the conditions that are considered conducive for such storms can be studied.

It has long been known that increased thermodynamic instability, increased moisture content in the atmosphere, and increased vertical wind shear within 5km above the ground create an environment more favorable for a tornado outbreak. 8,9 In particular, tornadoes are more likely to occur when both low stability (reflected in high values of "convective available potential energy" or "CAPE"), and high shear are present. Secondarily, the presence of an elevated mixed layer (reflected in moderate values of "convective inhibition" or "CIN") can delay the onset of storms, such that when they occur, they do so more explosively and in the form of more long-lived, isolated supercells, which can spawn tornadoes. A recent analysis of climate change projections suggests that the number of days during which meteorological conditions are favorable for severe storms may increase during latter decades of the 21st Century, primarily due to increased instability, though projected decreases in vertical wind shear may oppose thermodynamic destabilization.10

CAPE and PW 1980-2010
Fig. 2. 1979-2010 time series of April total column precipitable water (PW, top and second panel, kg/m2) and CAPE (third panel, J/kg), and the magnitude of the vector wind shear in the surface-500 millibar layer (bottom, m/s). Precipitable water is shown as averages for two regions; the Gulf of Mexico (top) and the lower Mississippi Valley (second panel). CAPE and shear time series are for the lower Mississippi Valley. The CAPE and shear are based on the single reanalysis (CFS-R). Also shown for the Gulf of Mexico region is the satellite-based estimate of total column water vapor for the period after 1987 (top, blue curve). Gulf of Mexico is defined as the area 20N-30N, 80W-95W. Lower Mississippi Valley defined as the area 31N-40N, 82W-95W.

Here we show a preliminary analysis, covering 1979-2010, to diagnose whether changes in such large-scale, time-averaged climate variables for April are detected**. Our diagnosis attempts to reveal whether large-scale conditions may have become more favorable for violent storms to occur over the lower Mississippi Valley, the region of the recent super tornado outbreak. Various data sets are used to estimate the time variability in column precipitable water (Fig. 2, top 2 panels). These are found to be in close agreement with each other with respect to their interannual variability; the water vapor time series are dominated by strong year-to-year variations both over the Gulf of Mexico (top) source region and over the lower Mississippi Valley impact region of possible tornadic activity (second panel). A similar interpretation applies to a time series of thermodynamic stability (based on the index of convective available potential energy, CAPE) which as might be expected, varies coherently with the atmospheric water vapor content (third panel). Finally, the vector wind shear magnitude for the surface-500mb layer is also dominated by interannual variability, with little evidence for a trend during the 30-yr period (lower panel).

Neither the time series of thermodynamic nor dynamic variables suggests the presence of a discernable trend for April conditions during the last 30 years; any small trend that may exist would be statistically insignificant relative to the intensity of yearly fluctuations. A change in the mean climate properties that are believed to be particularly relevant to major destructive tornado events has thus not been detected for April, at least during the last 30 years. So far, we have not been able to link any of the major causes of the tornado outbreak to global warming. Barring a detection of change, a claim of attribution (to human impacts) is thus problematic, although it does not exclude that a future change in such environmental conditions may occur as anthropogenic greenhouse gas forcing increases.10

The body of knowledge regarding the possible role played by large-scale climate forcings in tornado outbreaks is rapidly evolving, and constitutes a field of study that must integrate existing expertise in meso-scale meteorology with expertise in global-scale climate dynamics. Likewise, the methods for conducting attribution science also continue to evolve, and advances on the tornado-climate linkages will require modeling capabilities beyond current tools. Despite various limitations in data and tools, it should be noted that applying a scientific process is essential if one is to overcome the lack of rigor inherent in attribution claims that are all too often based on mere coincidental associations.

This assessment attempts to summarize a current scientific understanding of the link between climate and tornadoes. The assessment is not in lieu of a more rigorous diagnosis, nor does it preempt the need for better quantification of the physical climate factors associated with tornado outbreaks. This must be the topic of future research.

** An implicit assumption, which should be verified in future research, is that a change in the monthly average conditions would indicate similar changes in the tails of the daily distribution, i.e., a slight increase in the mean monthly CAPE would result in more individual days with CAPE sufficiently high to support tornadic activity. Similarly, future research would explore the frequency and change over time of the joint instantaneous occurrence of the high CAPE and high shear, and possibly moderate CIN that are conducive for tornado development.

Overview | El Niño/La Niña | Climate Change | Q & A | References