How fupit gets its numbers

Every value on fupit traces to real climate science. We do not invent coefficients or warming rates. Where we cannot ground a number, we leave it blank rather than guess.

Forecast sources

• Temperature and precipitation change: raw CMIP6 model output behind the IPCC Sixth Assessment Report, aggregated by scenario and decade.
• Present-day baseline: WorldClim v2.1 observed monthly climatology (1970-2000, 10 arc-minutes) where available, with CMIP6 historical climatology as fallback.
• Sea-level rise: IPCC AR6 regional projections.
• Heat, drought, and flood risk: CMIP6 ETCCDI extreme-climate indices scored against documented thresholds.
• Humid heat: NASA NEX-GDDP-CMIP6 daily-mean wet-bulb (Stull 2011) exceedance days above 28/31/35°C for the ~25 km cell; a regional screen, not measured WBGT.
• Cold-season context: NASA NEX-GDDP-CMIP6 daily ETCCDI indices (frost days, ice days, coldest-night TNn, cold-spell screen) for the ~25 km cell, alongside monthly-mean freeze months.
• Climate twin: nearest present-day city in the indexed catalog by standardized monthly temperature and precipitation distance from grounded model output.

Honesty rules

• Temperature is shown with raw CMIP6 model consensus as the headline; the IPCC-calibrated value, adjustment, and calibration factor are shown for comparison.
• SSP2-4.5 is the default middle-path reference; SSP5-8.5 is available as a very-high-emissions stress test, not as a business-as-usual claim.
• Precipitation is shown as model consensus plus spread because there is no equivalent single assessed calibration anchor.
• Risk scores expose the raw physical quantity next to the 0 to 100 score.

Sources: WorldClim v2.1 (Fick & Hijmans 2017); IPCC AR6 Working Group I; CMIP6; ETCCDI indices; Stull 2011 wet-bulb approximation; IPCC AR6 sea-level projections.

Return to fupit or inspect the source on GitHub.