Contents:

This repository contains an interactive map that presents detailed results from the very first application of the classification system, as in the journal article: The Development of an Adaptive Heat Stress Compensability Classification Applied to the United States, published in the 4th SNP special issue of the International Journal of Biometeorology. The results of this visualization were obtained from open-source data and coding packages such as Folium, and the model results were obtained by applying the Python Human Heat Balance (PyHHB) on a freely available weather data.

The interactive map provides a detailed visualization of the results for each city. When clicking on a city, a pie chart icon appears that contains the following three tabs:

Tab: Statistics: Details per city of the percentage of time in each category, as updated Figure 4b of the original publication.

Tab: Histogram 2D: Details per city as update of Figure 6 of the original publication. This histogram displays complementary avenues of human heat exchange on the X-axis and Y-axis to illustrate heat flux magnitudes in relation to compensable/uncompensable heat stress (UHS) and how UHS is disaggregated into the HSSC categories.

Tab: How to read the 2D histogram?: Details on how the categories are linked in the HSCC within 2D histogram areas (Figure 1 in the 2025 publication).

See the newest version: Changes in the HSCC interactive map

1. Updated the interactive supplemental map labels and subplots to account for the updates made in the IJB commentary paper.

From: Evaporative heat loss (W/m²) - Emax *ωmax + Eres  To: Evaporative heat loss (W/m²) - Min (Emax *ωmax , Emax sweat )+ Eres

Guzman-Echavarria, G., Middel, A., Vecellio, D.J., Vanos, J.K. (2025). The Heat Stress Compensability Classification (HSCC) applied to the United States: Update and Code, International Journal of Biometeorology.

2. Fixed an indexing bug for how we selected a single “peak hour” when Tmax values happened for more than one hour in a day. For those hot days that produced multi-hour peak hours, the code was computing averages instead of picking instantaneous values from a single hour. This is due to a pandas SettingWithCopyWarning that produced averaged values. The code was changed, and now it selects the first occurrence of the Tmax value and does explicit copy-and-loc assignments. This affects <0.05% of peak days, and we verify that there are no changes in the overall results, as you can see when comparing the current pie charts with the original publication.

For questions related to this dataset/code, please contact Gisel Guzman-Echavarria (gguzma20@asu.edu).

This application uses functions from: ASU/USYD HEAT-Lim (Human/Environmental Adaptation and Threshold Limits Model)
ASU/USYD Human/Environmental Adaptation and Threshold Limits Model (HEAT-Lim), and cite the original 2023 study (https://doi.org/10.1038/s41467-023-43121-5)  (https://zenodo.org/records/10020137).