Data for Diff-SPORT: Diffusion-based sensor placement optimization and reconstruction of turbulent flows in urban environments

Data for Diff-SPORT: Diffusion-based sensor placement optimization and
reconstruction of turbulent flows in urban environments
==============================================================================

Zenodo concept DOI: https://doi.org/10.5281/zenodo.22737508
Paper:    Vishwasrao et al., Nature Communications (in press); preprint arXiv:2506.00214
Code:     https://github.com/VinuesaLAB-AI/Diff-SPORT  (MIT license)
Contact:  Abhijeet Vishwasrao, University of Michigan (abvish@umich.edu)
License:  CC BY 4.0 for the files in this record, unless stated otherwise below.

This record holds the two datasets behind the paper: a direct numerical
simulation (DNS) of flow around a wall-mounted square cylinder, and an
experimental water-flume dataset of pollutant dispersion in an urban canopy.
Read the notes on provenance before using either one.

FILES IN THIS RECORD
-----------------------

DNS/DNS_data.zip
    DNS velocity data. Unzips to data/U.mat and data/V.mat (MATLAB v7.3,
    i.e. HDF5). 31 GB compressed, 33 GB uncompressed.

experimental/PURE_database.zip
    Verbatim copy of the public dataset "Dataset for Pollutant dispersion by
    tall buildings: Laboratory experiments and Large-Eddy Simulation",
    H. D. Lim et al., University of Southampton, DOI 10.5258/SOTON/D2217 

experimental/PURE_database_readme.txt
    The original Southampton cover readme, copied unchanged.

README.txt: Read for more details