Excess Heat & Cold Factors and Events

Global ERA5-Based Excess Apparent Temperature Dataset, 1940–2024 This repository contains a global daily dataset of absolute and seasonally relative excess apparent temperature events from 1940–2024. Datasets are first organized into annual-level .zip folders (one folder for every year, 1940 through 2024), and then within each annual folder, data are provided as monthly NetCDF files on a flattened ERA5-based grid with 1,038,240 spatial points. The dataset was produced by Cameron C. Lee, Ph.D., and the ClimRISE Laboratory at Kent State University. The source variable, AppTemp, is daily mean apparent temperature in degrees Celsius, calculated using the Steadman (1984) apparent temperature formulation in the shade. AppTemp is calculated from hourly ERA5 2m temperature, 2m dewpoint, 10m u-wind, and 10m v-wind. The 24 hourly apparent temperatures are calculated first, then averaged to produce daily mean AppTemp. Absolute and relative excess heat/cold metrics follow Sheridan and Lee (2018), based partly on Nairn and Fawcett (2015). Absolute metrics use fixed 1991–2020 reference periods (RFs)/thresholds: the 95th percentile for excess heat and the 5th percentile for excess cold. Relative metrics use seasonally varying 1991–2020 RFs/thresholds: the monthly 92.5th percentile for relative excess heat and monthly 7.5th percentile for relative excess cold, expanded into a full seasonal cycle and applied to 1940–2024. Absolute events generally occur in the expected season, while relative events are seasonal anomalies and can occur at any time of year, anywhere on Earth. Each NetCDF file includes AppTemp, excess heat/cold factors (EHF, ECF), relative factors (REHF, RECF), day indicators (EHday, ECday, REHday, RECday), and event indicators (EHEvents, ECEvents, REHEvents, RECEvents). Day and event indicators are binary uint8 values. AppTemp is packed int16 with scale_factor=0.01; factors are packed int16 with scale_factor=0.1. CF-aware software should unpack these automatically. Heat factors are positive and cold factors are negative. Event thresholds are based on the 85th percentile of nonzero heat factors and the 15th percentile of nonzero cold factors during 1991–2020, then applied to the full record. The first 32 days of factor fields are missing because acclimatization requires a 32-day lookback period; corresponding indicators are false. Files also include latitude, longitude, ERA5-based grid indices, and a seven-column time matrix containing MATLAB date number, year, month, day, hour, minute, and second. Further information can be found in the README file (.txt and/or .md), and the file_manifest (.csv). Feel free to use free of charge, but cite this repository and the appropriate development literature (some of which is listed below). Reach out to the lead author for further questions References: Ibebuchi, C.C., Lee, C.C. and Sheridan, S.C., 2025. Recent trends in extreme temperature events across the contiguous United States. International Journal of Climatology, 45(2), p.e8693. Ibebuchi, C.C., Lee, C.C. and Sheridan, S.C., 2026. Risk assessment of counties in the contiguous United States impacted by increasing frequency of hazardous temperatures. Natural Hazards, 122(1), p.14. Nairn, J.R. and Fawcett, R.J., 2015. The excess heat factor: a metric for heatwave intensity and its use in classifying heatwave severity. International Journal of Environmental Research and Public Health, 12(1), pp.227–253. Sheridan, S.C. and Lee, C.C., 2018. Temporal trends in absolute and relative extreme temperature events across North America. Journal of Geophysical Research: Atmospheres, 123(21), pp.11889–11908. Steadman, R.G., 1984. A universal scale of apparent temperature. Journal of Applied Meteorology and Climatology, 23(12), pp.1674–1687.