Add GeoRank city climate normals dataset - #11
stoffis-git wants to merge 2 commits into
Conversation
|
Small update since opening this: the dataset is now archived with a permanent DOI at Zenodo (CERN), so the entry points at something citable and version-stable rather than only a live URL. https://doi.org/10.5281/zenodo.21976372 — CC BY 4.0, with the method and the full 219-station reference list linked from the record. Happy to amend the entry to cite the DOI instead of (or alongside) the URL if that fits the list better — just say which you prefer and I will push it. No rush, and no problem at all if the entry is not a fit. |
|
Trimmed this to fit the list. The entry was 124 characters of link text, where the longest entry on the list is 79 and the median is 39 — it read as an advert rather than an index entry. It is now 66 characters and in the same shape as the neighbouring lines. Also now archived with a permanent DOI at Zenodo: https://doi.org/10.5281/zenodo.21976372 — happy to point the entry there instead if you would prefer a citable archive over the live URL. |
Adds one entry under Data Sets → Single Application.
What it is: derived climate normals for 478 cities — monthly sunshine, mean and daytime-high temperature, rainfall, and an outdoor-comfort measure that counts daylight hours lost to heat, cold, rain and wind separately.
Against the scope in CONTRIBUTING.md:
Provenance, documented at https://georank.place/methodology: CHELSA v2.1 temperature and precipitation, ERA5 humidity and wind, and sunshine bias-corrected against 219 WMO, KNMI, ECA&D and national met-service reference stations, each listed individually with its source and normals period.
Relevance to the list's mission: the heat/cold attribution makes habitability impacts measurable per city rather than as a global average — for example the daylight hours a city already loses to heat in a typical year, which is the observed counterpart to projected heat-stress work.
Disclosure: I built this, so treat the suggestion accordingly. Two honest limitations: the underlying normals are 1981–2010 (sunshine 1991–2020), so it describes a recent-historical baseline rather than current or projected conditions, and the outdoor-comfort thresholds are ours rather than an official standard — the per-city ranking is far more robust than any absolute figure. Happy to reword, move it to a different section, or close this if it is not a fit.