Dataset: Urban thermal hotspots through integration of geoparticipation and high-resolution modelling

README General information The dataset serves as supplementary data for a paper titled 'Urban thermal hotspots through integration of geoparticipation and high-resolution modelling'. Note that the manuscript has not been reviewed yet. Dataset description The archive includes the following: 1. Folder 00_Calibration-measurements contains calibration measurement data for the PALM model. The input data required for a simulation run exceeds the repository size (~75 GB). If someone is interested in static and dynamic drivers, please contact Dr Geletič. 2. Folder 01_Paths contains original paths of individual respondents. Basic description is in file Holesovice2024_popis.docx. Note that all data were anonymised. 3. Folder 02_TSV contains merged anonymous TSV locations, merge_shp_heatwalk.shp, for all respondents (see respondents' ID in attribute table or in table tab_HOLE_heatwalk.csv). Corrected and preprocessed locations are stored in the subfolder 'individual'. 4. Folder 03_PALM-results contains post-processed simulation results. The original NetCDF file is too large (~145 GB), so we share the exported GeoTiff files for the experimental period and the selected variables only. Each file (GeoTiff, or *.prj) has the same naming convention. As an example, bio_utci_abs-01min_20240806_1300-1301.tif, could be parsed as: variable name (bio_mrt), type of the processed output - absolute 'abs' value during the 1 minute (abs-01min), date (20240806) and averaged period (1300-1301) in local time (CEST, +2 UTC). Folder contains three subfolders: rst_UTCI universal thermal climate index (UTCI). PALM variable name is 'bio_utci*_xy', but was renamed to 'bio_utci'. Unit is degree of Celsious. rst_T2m with surface air temperature in 2m. PALM variable name is 'theta_2m*_xy', but was renamed to 'theta_2m'. Unit is degree of Celsious. rst_MRT with mean radiant temperature (MRT). PALM variable name is 'bio_mrt*_xy', but was renamed to 'bio_mrt'. Unit is degree of Celsious. Abstract Ongoing urbanisation and climate change impacts in cities require increasingly accurate, spatially explicit data to address deteriorating human outdoor thermal comfort. Globally, urban heat exposure threatens billions of people, yet cities lack tools to identify where residents actually experience thermal stress and where interventions are most needed. We developed an integrated approach combining citizen-science thermal walks (n = 3,666) with high-fidelity microclimate modelling to map a thermal hotspot in Holešovice, a district of Prague, the capital of the Czech Republic. The results confirm a significant relationship among thermal sensation votes (TSV), mean radiant temperature (MRT), and the Universal Thermal Climate Index (UTCI). In contrast, surface air temperature has a substantially different spatial pattern. By understanding the role of heat stress mitigation measures, local municipalities can maximise their impact in vulnerable areas. This precision approach enables evidence-based adaptation planning and targets resources where they are most needed. With emerging AI technologies, our methodology offers scalable solutions for resource-constrained cities globally. ACKNOWLEDGEMENT We thank all participants who voluntarily took part in the thermal walk and contributed their thermal perception data to this study. We acknowledge the Ministry of Education, Youth and Sports of the Czech Republic for providing HPC resources through the e-INFRA CZ (ID:90254). The field campaign was supported by the Strategy AV21 project Materiality (Czech Academy of Sciences). PALM model development was supported by the Johannes Amos Comenius Programme (oP JAC), project No. CZ.02.01.01/00/22_008/0004605. The dataset is published under the Creative Commons Attribution 4.0 International License (CC-BY-NC-4.0). This license allows others to distribute, remix, adapt, and build upon the dataset for any purpose, even commercially, as long as they give appropriate credit to the original creator(s).