Tierney et al. (2025) A novel framework for expanding temperature intensity-duration-frequency curve utility

Identification

• Journal: Natural Hazards
• Year: 2025
• Date: 2025-12-22
• Authors: Gregory Tierney, Megan S. Mallard, Tanya L. Spero, Anna M. Jalowska
• DOI: 10.1007/s11069-025-07811-1

Research Groups

• State Climate Office of North Carolina, North Carolina State University, Raleigh, NC, USA
• U.S. Environmental Protection Agency, Research Triangle Park, NC, USA

Short Summary

This study presents an expanded framework for temperature intensity-duration-frequency (TIDF) curve analysis, including an objective fitting algorithm and uncertainty quantification, to assess the risk of extreme hot, cold, and "near-extreme" temperature events. The framework demonstrates its utility by contextualizing recent record-breaking heat waves and cold snaps, providing a flexible and robust tool for hazard response and adaptation planning.

Objective

• To develop and present an expanded framework for temperature intensity-duration-frequency (TIDF) curve analysis, incorporating an objective fitting algorithm and uncertainty quantification for observational records.
• To extend the application of TIDF curves beyond heat waves to include extreme cold events and "near-extreme" events (e.g., 5th and 95th percentiles) for a more comprehensive quantification of extreme temperature risk.
• To calculate and communicate confidence intervals around TIDF curves to provide additional context for the severity and unprecedented nature of historical extreme events.

Study Configuration

• Spatial Scale: Three individual meteorological stations in the United States: Raleigh-Durham International Airport, North Carolina (KRDU); Seattle-Tacoma International Airport, Washington (KSEA); and Dallas/Fort Worth International Airport, Texas (KDFW).
• Temporal Scale:
  • Baseline analyses: 30 water years (1 October 1990 to 30 September 2020).
  • Event durations: 13 durations considered, including sub-daily (2, 6, 12 hours) and daily (24 to 240 hours, stepped every 24 hours).
  • Case studies: 2021 Pacific Northwest heat wave (June 2021) and 2021 Texas cold snap (February 2021), analyzed against antecedent records (1990–2020) and extended baselines (e.g., 1980–2020 for KDFW).

Methodology and Data

• Models used:
  • Statistical distributions for fitting Annual Percentile Series (APS): cosine, hyperbolic secant, logistic, and normal. A uniform distribution was used as a data quality flag.
  • Objective algorithm for distribution selection based on maximum likelihood estimation and Kolmogorov–Smirnov (K–S) test statistics.
  • Bootstrapping method (7500 resamples) for calculating 90% confidence intervals (CIs).
• Data sources: Hourly temperature observations from the Integrated Surface Dataset (ISD) distributed by the National Centers for Environmental Information (NCEI).

Main Results

• The expanded TIDF framework successfully analyzes extreme hot (100th percentile), extreme cold (0th percentile), and "near-extreme" (5th and 95th percentiles) temperature events.
• TIDF curves exhibit monotonic behavior with an abrupt slope change between sub-daily (2, 6, 12 hours) and daily (≥24 hours) event durations, reflecting the influence of overnight temperatures.
• Confidence intervals (90% CI) for TIDF curves are generally wider for cold extremes than for hot extremes, indicating lower confidence in the range of cold extreme event intensities.
• For non-absolute extremes, uncertainty (CI width) typically ranges from 1.1 °C to 3.3 °C across different stations, percentiles, and distributions.
• 2021 Pacific Northwest heat wave (KSEA):
  • The event was estimated as a 1-in-500-year event relative to the 1990–2020 baseline.
  • Annual maxima for 2020–2021 exceeded the 1-in-100-year TIDF curve for durations up to 120 hours (5 days) but remained within the 90% CI.
  • Including the event in the baseline (1990–2021) increased the 1-in-100-year median event by an average of 0.53 °C and its CI width by 13% (from 4.2 °C to 4.8 °C).
• 2021 Texas cold snap (KDFW):
  • The event was estimated as a 1-in-750-year event at the shortest durations and 1-in-100-year to 1-in-500-year for most other durations relative to the 1990–2020 baseline.
  • Annual minima for 2020–2021 exceeded the 1-in-100-year threshold across all durations (2 hours to 10 days), with significant exceedances at sub-daily and 8-9 day durations.
  • The event largely fell within the 90% CI of a 1-in-100-year event, and lengthening the baseline to 1980–2020 changed its rarity estimate to 1-in-50-year to 1-in-100-year at median thresholds, underscoring the importance of CIs.

Contributions

• Generalizes the TIDF curve concept by utilizing Annual Percentile Series (APS), enabling the analysis of both extreme hot and cold events, as well as "near-extreme" events (e.g., 5th and 95th percentiles), for a more robust and comprehensive risk assessment.
• Introduces an objective algorithm for selecting the best statistical distribution fit for TIDF applications, enhancing transparency, adaptability, and automation compared to previous subjective methods.
• Integrates uncertainty quantification through bootstrapping-derived confidence intervals (CIs) into TIDF analysis, providing stakeholders with a more complete understanding of potential event intensities and their associated uncertainties.
• Demonstrates the practical utility and cross-sector applicability of the TIDF framework by effectively contextualizing two recent, high-impact extreme temperature events (2021 Pacific Northwest heat wave and 2021 Texas cold snap).
• Offers a consistent, data-driven methodology for evaluating extreme temperature risk that is flexible across different locations and planning needs, without relying on predefined, potentially arbitrary, temperature thresholds.

Funding

• U.S. Environmental Protection Agency (EPA) Research Participation Program administered by the Oak Ridge Institute for Science and Education (ORISE).
• Interagency agreement between the U.S. Department of Energy (DOE) and the U.S. Environmental Protection Agency.
• ORISE managed by ORAU under DOE contract number DE-SC0014664.

Citation

@article{Tierney2025novel,
  author = {Tierney, Gregory and Mallard, Megan S. and Spero, Tanya L. and Jalowska, Anna M.},
  title = {A novel framework for expanding temperature intensity-duration-frequency curve utility},
  journal = {Natural Hazards},
  year = {2025},
  doi = {10.1007/s11069-025-07811-1},
  url = {https://doi.org/10.1007/s11069-025-07811-1}
}