Citation:

Tierney, G., M. Mallard, G. Gray, A. Jalowska, T. Spero, AND J. Bowden. Expanding the Utility of Intensity-Duration-Frequency (IDF) Curves for Extreme Temperature Events. American Meteorological Society's 103rd Annual Meeting, Denver, CO, January 08 - 12, 2023.

Impact/Purpose:

Extreme temperature events -- both heat waves and cold snaps -- are some of the most high-impact weather events across the country. They present a continuing danger to human health, especially in underserved communities where indoor climate control is not as prevalent and outdoor laborers make up a significant portion of the workforce. Moreover, these events often require a community response to activate indoor shelters to help mitigate negative health outcomes -- a crucial step to protect vulnerable populations. Recent events, such as the 2021 heat wave in the Pacific Northwest, have further reinforced this reality, and serve as a warning that such events may have even greater impact if the Earth's climate warms as projected. This work demonstrates the flexibility of intensity-duration-frequency (IDF) curves in evaluating the current and future occurrence and strength of extreme temperature events. Our results feature compact visualization of extreme temperature event potential at several durations (from 2 hours to 10 days) and several return periods (1-in-2 year to 1-in-100 year events), all in a single chart. The capacity to also provide confidence intervals is demonstrated, quantifying the variability in the estimates. Finally, several applications are showcased, including the use of human health metrics such as heat index and integration with National Weather Service forecasts and observations -- providing stakeholders with real-time context for extreme events. This work has potential to be used for long-term (multi-decade) and short-term (days to weeks) planning. Long term applications include positioning communities and partners to be better prepared for what the strength and frequency of future extreme temperature events may be (and thus the associated emergency response that will be required) by combining the IDF analysis with downscaled climate projections. Short term applications include providing real-time context for events in the immediate future, allowing stakeholders to respond and allocate emergency resources effectively.

Description:

Extreme temperatures – both heat waves and cold events – pose a great hazard to human health, property, and infrastructure. Between the years 1998 and 2017, the World Health Organization estimates that on average, more than 8,300 people worldwide died each year due to heat waves, with losses disproportionately affecting underserved and vulnerable communities. Among other physiological factors, both hot and cold extreme events increase cardiovascular stress within the human body, contributing to increases in mortality due to increased cardiovascular load. Furthermore, these events are projected to become more common and more extreme as the Earth’s climate system warms. This complicates adaptation and mitigation planning for local governments providing cooling/warming centers, companies depending on outdoor labor, and utilities anticipating changing energy demands, along with many other stakeholders. Flexible tools are needed to evaluate the current and future risk of extreme temperature events spanning multiple durations or thresholds with the potential to affect human health and the environment.                 To that end, we present an expanded framework for using intensity-duration-frequency (IDF) curves to evaluate temperature extremes. While IDF curves are commonly used in the precipitation modeling and hydrological communities, the application of IDF curves to temperature extremes has been more limited and focused on heat waves exclusively. Here, we expand that application to consider their utility for cold events as well as other “near-extreme” scenarios, which can be similarly impactful but are often overshadowed by the most extreme events. The information-dense format of an IDF curve permits the simultaneous consideration of multiple return periods (from multi-year to multi-decade) and event lengths (from sub-daily durations to beyond 7 days), while also remaining visually straightforward to interpret. Using a probabilistic approach to extremes, we also calculate confidence intervals for all return periods and durations, providing additional context for the severity and unprecedented nature of historically extreme events. Further analysis of confidence interval width provides further insight into differences between hot and cold extremes, as well as the potential impact of different sampling methods when considering how these potentially devastating events might change in the future.

URLs/Downloads:

Record Details: