GSOP Climate Observatory Prototype
Evidence for People. Data for Action.
Evidence for People. Data for Action.
Explore the evidence cell by cell across the globe.
Explore NOAA weather-station records, regional coverage, and annual temperature charts. Compare three periods with experimental trends; the records are not homogenized and coverage varies by place.
The regional surface and clickable 0.5° map use the published NOAA/PSL GHCN-CAMS land-grid release. NOAA GSOY station trends and inventory counts are supplementary evidence; the station counts below are not CAMS land-area coverage and are not a planet-wide percentage.
The default view uses NOAA GSOY annual station observations. Each station independently needs at least 32 years in 1965–2004, 17 in 2005–2025, or 32 in 1986–2025. Cell trends use separate station intercepts (qualified_station_fixed_effects_v1), so constant temperature differences between stations do not create a trend when station membership changes. These experimental estimates are not homogenized; station moves, instrument changes and uncertainty remain unresolved. The separate NOAA analyzed comparison uses the published NOAA/PSL GHCN-CAMS field.
…/access/{station_id}.csv).The default public map uses qualifying NOAA annual station estimates and separate qualifying NOAA monthly-derived estimates on the shared 0.5° grid. Annual station estimates take priority over qualifying monthly-derived estimates at the same cell and period. Tom’s Original GSOP archive is retained for discovery, provenance review, and its explicit archive view; it does not fill public station-derived cells. Sources are never averaged together or spliced. At overview zoom, consistent 5° display cells summarize station-derived evidence; archive summaries are not used in the public station-derived surface. Sparse evidence remains marked separately. Zoom in for the shared 0.5° station-derived cells and source-specific evidence. Original GSOP retains its published 1965–2004 and 2005–2025 aggregate results only in the explicit archive view.
Uncolored cells have no qualifying snapshot estimate. Regional evidence shows the qualifying count out of 100 cells, and every cell can be opened to inspect live records or a no-result explanation. Coverage is uneven and changes with the selected period; counts do not measure global land-area coverage. A missing estimate is not zero warming. Differences between period estimates alone do not establish acceleration.
This is a demonstration, not a final peer-reviewed scientific product. The interface reports the period, usable years and contributing station count for each computed trend.