9.4 ENSO-driven Seasonal Variability in Hail, Tornadoes and Losses

Wednesday, 24 October 2018: 12:00 PM
Pinnacle room (Stoweflake Mountain Resort )
John T. Allen, Central Michigan Univ., Mount Pleasant, MI; and M. J. Molina, V. A. Gensini, E. Faust, M. Steuer, and J. Eichner

The likelihood of United States (U.S.) seasonal hail and tornado occurrence have been previously shown to be related to the variability of Pacific Sea-Surface Temperatures associated with the El Niño-Southern Oscillation (ENSO). However, these earlier works assume the recognized stationary framework in which tornado season peaks in late May. Long-term trends in tornado observations suggest a shift toward earlier tornado season peaks, and yet fail to examine the role of year-to-year climate variability. Exploring the influence of ENSO on daily probabilities, modulation of annual cycle shape for U.S. tornadoes is demonstrated, in addition to modification of the overall probability of occurrence. Observations and favorable environments show substantial modification of the peak spatial distribution and the temporal onset of tornado occurrence. La Niña produces an earlier annual peak probability by 1.5–2 weeks, with a higher overall fraction of events in March and April. In contrast, El Niño leads to a week delay in the maximum probability, and enhances a second peak in the fall months. Consequently, this suggests that climate change is not the sole driver of changes to seasonal onset and peak, and that climate variability plays an important role in modulating the annual cycle.

Yet for all the discussion of the physical influence of ENSO on tornado and hail occurrence, do we see a response in U.S. insured and uninsured losses associated with severe thunderstorms? Based on the data and loss estimation method used in the Munich Re NatCatSERVICE database, a dataset of U.S. severe convective storm losses was created. The direct overall and insured losses available on a state level since 1970 were normalized to today’s level of destructible assets in order to remove the signal of economic growth from the losses. Such preprocessing of the data preconditions losses to reflect the impact of hazard conditions varying due to different ENSO phases. This presentation will address the question of the impacts of ENSO on loss, and discuss what we know about the role of ENSO on severe weather, its limitations for predictability, and briefly cover the failure modes for the 2018 spring seasonal forecast.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner