WeatherPerch report · updates continuously
Weather anomaly index
The trailing 7 days of forecast data across every WeatherPerch-tracked location, scored against each location's own 30-day trailing baseline. Nothing below is written prose — every row is a z-score comparison against real observed history, not an editorial judgment call.
This week's anomalies
| Score | Location | Date | Type | Signal |
|---|---|---|---|---|
| 93 | Srinagar, India | 2026-10-04 | Temperature swing | A wide temperature swing is expected on 2026-10-04. |
| 92 | Srinagar, India | 2026-10-01 | Temperature swing | A wide temperature swing is expected on 2026-10-01. |
| 92 | Little Rock, United States | 2026-09-30 | Temperature swing | A wide temperature swing is expected on 2026-09-30. |
| 91 | Srinagar, India | 2026-10-05 | Temperature swing | A wide temperature swing is expected on 2026-10-05. |
| 91 | Srinagar, India | 2026-10-03 | Temperature swing | A wide temperature swing is expected on 2026-10-03. |
| 90 | Srinagar, India | 2026-09-30 | Temperature swing | A wide temperature swing is expected on 2026-09-30. |
| 90 | Tulsa, United States | 2026-09-28 | Temperature swing | A wide temperature swing is expected on 2026-09-28. |
| 89 | Srinagar, India | 2026-10-02 | Temperature swing | A wide temperature swing is expected on 2026-10-02. |
| 89 | Brasília, Brazil | 2026-09-30 | Temperature swing | A wide temperature swing is expected on 2026-09-30. |
| 89 | Srinagar, India | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 87 | Queenstown, New Zealand | 2026-09-30 | Temperature swing | A wide temperature swing is expected on 2026-09-30. |
| 87 | Nashville, United States | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 87 | Srinagar, India | 2026-09-28 | Temperature swing | A wide temperature swing is expected on 2026-09-28. |
| 86 | Oxford, United Kingdom | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 86 | Coventry, United Kingdom | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 85 | Panāji Muwara, India | 2026-10-02 | Temperature swing | A wide temperature swing is expected on 2026-10-02. |
| 85 | Nashville, United States | 2026-09-30 | Temperature swing | A wide temperature swing is expected on 2026-09-30. |
| 84 | Tureni, Romania | 2026-10-04 | Temperature swing | A wide temperature swing is expected on 2026-10-04. |
| 84 | Tel Aviv, Israel | 2026-10-04 | Temperature swing | A wide temperature swing is expected on 2026-10-04. |
| 84 | Vadodara, India | 2026-10-03 | Temperature swing | A wide temperature swing is expected on 2026-10-03. |
| 84 | Memphis, United States | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 84 | Bradford, United Kingdom | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 84 | Sheffield, United Kingdom | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 84 | Leeds, United Kingdom | 2026-09-29 | Temperature swing | A wide temperature swing is expected on 2026-09-29. |
| 84 | Dubai, United Arab Emirates | 2026-09-28 | Temperature swing | A wide temperature swing is expected on 2026-09-28. |
How this is scored
Each location's daily rainfall, temperature range, and wind speed are compared to that same location's own trailing 30-day mean and standard deviation — a WMO SPI-style z-score, not a fixed global threshold. A score only qualifies once it clears both a real-world minimum (e.g. 10"+ of rain, not just "more than usual" for a desert) and a statistical rarity bar (z ≥ 1.5, roughly the top 7% of days for that location). Forecast-shift events instead measure how much a location's own forecast changed between consecutive ingestion runs.
Only locations with engine version 2 results are shown, so a scoring-methodology change never silently mixes old and new results in the same table. A short trailing history (fewer than 3 samples) falls back to a conservative fraction-of-mean estimate rather than a divide-by-near-zero spike.
This index is descriptive, not predictive — it reports what the data already shows, not what will happen. Official NWS alerts remain the authority for safety decisions; see the live alert banner on any location's forecast page.
Download this report (JSON) → · Read the full methodology →
Want this on your own site? Embed the anomaly index →