---
title: Storm Charles through the chain
summary: The same storm as each layer of the weather supply chain saw it: stations and the airport, the models, the aviation forecast, IPMA's warnings, MeteoAlarm, and the sites people look at. From the data room only.
mount: <div class="card"><p class="kicker">One storm, six layers</p><div class="wx-tiles"><article class="wx-tile "><p class="wx-th"><span class="lbl l-fact">fact</span> <b>1. Measured</b></p><p class="wx-big">67 km/h<small>strongest gust so far at Lisbon airport (11:00 Tue); 249.5 mm of rain in 6 hours at Madeira, Areeiro</small></p><p class="wx-src"><span class="prov" style="white-space:normal"><a href="https://aviationweather.gov/api/data/metar?ids=LPPT&amp;hours=30&amp;format=json">METAR via NOAA AWC</a>, read <time datetime="2026-11-11T11:26:40Z">11:26 UTC, 11 Nov</time></span> <span class="prov" style="white-space:normal"><a href="https://api.ipma.pt/open-data/observation/meteorology/stations/observations.json">IPMA stations</a>, read <time datetime="2026-11-11T11:26:40Z">11:26 UTC, 11 Nov</time></span></p></article><article class="wx-tile "><p class="wx-th"><span class="lbl l-fact">fact</span> <b>2. Computed by a model</b></p><p class="wx-big">40 km/h<small>strongest mean wind on Wednesday in IPMA's hourly model forecast for Lisbon (14:00 Wed): a mean for a grid cell, not a gust</small></p><p class="wx-src"><span class="prov" style="white-space:normal"><a href="https://api.ipma.pt/public-data/forecast/aggregate/1110600.json">IPMA hourly forecast</a>, read <time datetime="2026-11-11T11:26:40Z">11:26 UTC, 11 Nov</time></span></p></article><article class="wx-tile "><p class="wx-th"><span class="lbl l-fact">fact</span> <b>3. Forecast for pilots (TAF)</b></p><p class="wx-big">102 km/h<small>strongest gust group for Lisbon airport, probability 30%, 12:00 Wed to 16:00 Wed</small></p><p class="wx-src"><span class="prov" style="white-space:normal"><a href="https://aviationweather.gov/api/data/taf?ids=LPPT&amp;format=json">TAF via NOAA AWC</a>, read <time datetime="2026-11-11T11:26:40Z">11:26 UTC, 11 Nov</time></span></p></article><article class="wx-tile t-red"><p class="wx-th"><span class="lbl l-fact">fact</span> <b>4. Warned by IPMA</b></p><p class="wx-big"><span class="lvl red">red</span> rain<small><span lang="pt">Precipitação forte e persistente.</span> 09:00 Wed to 18:00 Wed</small></p><p class="wx-src"><span class="prov" style="white-space:normal"><a href="https://api.ipma.pt/open-data/forecast/warnings/warnings_www.json">IPMA warnings</a>, read <time datetime="2026-11-11T11:26:40Z">11:26 UTC, 11 Nov</time></span></p></article><article class="wx-tile "><p class="wx-th"><span class="lbl l-fact">fact</span> <b>5. Carried by MeteoAlarm</b></p><p class="wx-big">18 messages<small>IPMA's warnings for Lisbon as CAP, each with the time IPMA sent it</small></p><p class="wx-src"><span class="prov" style="white-space:normal"><a href="https://feeds.meteoalarm.org/api/v1/warnings/feeds-portugal">MeteoAlarm feed</a>, read <time datetime="2026-11-11T11:26:40Z">11:26 UTC, 11 Nov</time></span></p></article><article class="wx-tile "><p class="wx-th"><span class="lbl l-evidence">evidence</span> <b>6. Drawn by sites</b></p><p class="wx-big">layer 2, mostly<small>the automatic wind on a weather page is a model's mean for a grid cell; the warning is a forecaster's decision about the worst gusts</small></p><p class="wx-src"><span class="prov"><a href="../research/weather-sites/index.html">incident.weatherwatch@'s reading</a></span></p></article></div><p class="meta">The same storm at the same hours, four different kinds of number. None contradicts another: each layer answers a different question. The warning is layer 4; nothing here is a forecast of ours.</p></div>
---
# Storm Charles through the chain

> Kept by incident.meteorology@web-summit.dinis, 11 November 2026, 11:26 UTC

[Where a forecast comes from](index.md) · from [the data room's weather files](../incidents/2026-11-10-storm-charles/data/weather/README.md) only, as they stood at the reads given. Not a forecast of ours and not a warning: **the warnings are IPMA's** (layer 4). Each layer below is labelled: a **fact** was read from the body that measures, decides or warns; **evidence** is what others computed or reported; nothing here is our own prediction. Times are UTC (Lisbon is on UTC in November).

## 1. What was measured (observing systems)

**Fact.** [IPMA's surface stations](nodes/obs-ipma-stations.md): eight stations in and next to Lisbon, hourly means, **no gusts in this feed**. Strongest hourly mean wind kept so far at each ([IPMA observations](https://api.ipma.pt/open-data/observation/meteorology/stations/observations.json), read 11:26 UTC, 11 Nov) · [stations.json](../api/v1/2026-11-10-storm-charles/weather/stations.json):

| Station | Strongest hourly mean wind | At | Latest hour: wind, rain, pressure |
|---|---|---|---|
| Lisboa (Geofísico) | 35.3 km/h | 05:00 Tue | 10:00 Wed: 13.7 km/h SE, ? mm, 1000.3 hPa |
| Lisboa, Relógio | 42.8 km/h | 04:00 Tue | 10:00 Wed: 16.2 km/h SE, 5.4 mm, no hPa |
| Lisboa, Tapada da Ajuda | 23.8 km/h | 13:00 Tue | 10:00 Wed: 9 km/h SE, 4.8 mm, no hPa |
| Lisboa, Amoreiras (LFCL) | 25.2 km/h | 04:00 Tue | 10:00 Wed: 10.8 km/h S, 3.8 mm, no hPa |
| Lisboa, Carnide (CML) | 29.9 km/h | 11:00 Tue | 10:00 Wed: 14.4 km/h S, 4.4 mm, 999.8 hPa |
| Oeiras / Vila Fria (CMO) | 37.1 km/h | 11:00 Tue | 10:00 Wed: 15.5 km/h SE, 5.9 mm, 999.7 hPa |
| Almada, P.Rainha | 24.5 km/h | 11:00 Tue | 10:00 Wed: 16.9 km/h SE, 5.6 mm, no hPa |
| Barreiro, Lavradio | 24.8 km/h | 23:00 Mon | 10:00 Wed: 18.4 km/h SE, 4.7 mm, no hPa |

**Fact.** [Airport observations (METAR)](nodes/obs-airport-metar.md), Lisbon airport (LPPT), the only gust series we hold: 92 reports kept; strongest gust so far **67 km/h** (36 kt) at 11:00 Tue ([NOAA Aviation Weather Center](https://aviationweather.gov/api/data/metar?ids=LPPT&hours=30&format=json), read 11:26 UTC, 11 Nov) · [airport.json](../api/v1/2026-11-10-storm-charles/weather/airport.json). The reports are issued in Portugal and redistributed by [NOAA Aviation Weather Center data API](nodes/data-awc-api.md): public domain as redistributed.

**Fact.** Madeira, IPMA's stations: the largest 6-hour rain at each ([IPMA observations](https://api.ipma.pt/open-data/observation/meteorology/stations/observations.json), read 11:26 UTC, 11 Nov) · [madeira.json](../api/v1/2026-11-10-storm-charles/weather/madeira.json):

| Station | Largest 6-hour rain | Hours (as IPMA labels them) |
|---|---|---|
| Madeira, Areeiro | **249.5 mm** | 01:00 Wed to 06:00 Wed |
| Madeira, São Vicente | **141.5 mm** | 22:00 Tue to 03:00 Wed |
| Madeira, Santo da Serra | **130.4 mm** | 21:00 Tue to 02:00 Wed |
| Madeira, Pico Alto | **126.8 mm** | 03:00 Wed to 08:00 Wed |
| Madeira, Pico do Areeiro | **123.9 mm** | 20:00 Tue to 01:00 Wed |

IPMA does not document whether an hour's label is the hour starting or ending; the data room's inference (hour ending) is labelled there.

Not in our data room: radar, satellite, radiosondes, buoys. They exist ([IPMA's weather radars](nodes/obs-radar-ipma.md), [Meteosat (geostationary)](nodes/obs-meteosat.md), [Radiosondes (weather balloons)](nodes/obs-radiosondes.md), [Buoys and ships](nodes/obs-buoys-ships.md)) and feed the models below; their licences decide whether we may show them ([licences](licences.md)).

## 2. What the models computed

**Fact (a model's output, published by IPMA).** IPMA's hourly forecast for Lisbon ([IPMA](https://api.ipma.pt/public-data/forecast/aggregate/1110600.json), read 11:26 UTC, 11 Nov) · [forecast.json](../api/v1/2026-11-10-storm-charles/weather/forecast.json): strongest **mean** wind on Wednesday **40 km/h** S at 14:00 Wed. IPMA does not say on that feed which model it is; IPMA's own pages say it runs [AROME (Météo-France)](nodes/model-arome.md) and derives products from [ECMWF IFS](nodes/model-ecmwf-ifs.md) (see [IPMA](nodes/body-ipma.md)).

The same feed, earlier runs kept by the data room (the forecast changed under the same stamp):

| Read | Wednesday's strongest mean wind | At |
|---|---|---|
| 20:22 UTC, 10 Nov | 48.6 km/h NW | 18:00 Wed |
| 00:00 UTC, 11 Nov | 23.4 km/h NW | 20:00 Wed |
| 08:22 UTC, 11 Nov | 23.4 km/h NW | 20:00 Wed |

**Evidence (other bodies' models, never merged with IPMA's).**

| Source | Model behind it | Wednesday's strongest mean wind | Strongest gust | Licence | Read |
|---|---|---|---|---|---|
| [MET Norway Locationforecast 2.0 (complete)](https://api.met.no/weatherapi/locationforecast/2.0/complete?lat=38.72&lon=-9.14) | [ECMWF IFS](nodes/model-ecmwf-ifs.md) (outside the Nordic area, per MET Norway) | 55.1 km/h at 15:00 Wed | not in this product | NLOD 2.0 / CC BY 4.0, data from MET Norway | 11:26 UTC, 11 Nov |
| [Open-Meteo (best match)](https://api.open-meteo.com/v1/forecast?latitude=38.72&longitude=-9.14&hourly=wind_speed_10m,wind_gusts_10m,wind_direction_10m,precipitation&timezone=UTC&start_date=2026-11-11&end_date=2026-11-12) | its "best match" mix: [Open-Meteo](nodes/data-open-meteo.md) | 18.9 km/h at 10:00 Wed | 50.8 km/h at 11:00 Wed | CC BY 4.0, Open-Meteo.com | 11:26 UTC, 11 Nov |

### Six models side by side, after the official statement

**The official statement first (fact).** IPMA's warnings in force or still to come on Wednesday 11 November ([IPMA](https://api.ipma.pt/open-data/forecast/warnings/warnings_www.json), read 11:26 UTC, 11 Nov) · [warnings.json](../api/v1/2026-11-10-storm-charles/weather/warnings.json):

- **Lisbon district:** **red** rain, 09:00 Wed to 18:00 Wed: "Precipitação forte e persistente."; **orange** rough sea, 12:00 Wed to 21:00 Wed: "Ondas de sudoeste com 5 a 7 metros e que poderão atingir até 13 metros de altura máxima."; **orange** wind, 09:00 Wed to 21:00 Wed: "Rajadas até 120 km/h."; **yellow** rough sea, 21:00 Wed to 03:00 Thu: "Ondas de sudoeste com 4 a 5 metros.".
- **Madeira (red and orange only in our file):** Madeira north coast: **orange** wind, 07:33 Wed to 12:00 Wed: "Rajadas até 100 km/h, em esepcial nos extremos leste e oeste."; Madeira south coast: **orange** wind, 07:33 Wed to 12:00 Wed: "Rajadas até 100 km/h."; Porto Santo: **orange** wind, 07:33 Wed to 12:00 Wed: "Rajadas até 100 km/h."; Madeira mountains: **orange** wind, 07:33 Wed to 12:00 Wed: "Rajadas até 130 km/h.".

These warnings are the statement that counts. What follows is not a forecast and not ours.

**Evidence.** Each row is a model's output, read at [Open-Meteo](nodes/data-open-meteo.md) ([our sample, as returned](samples/open-meteo-lisbon-funchal.json), read 02:49 UTC, 11 Nov), for the grid cell nearest each place; attribution: Weather data by Open-Meteo.com, CC BY 4.0. The figure is the highest gust value in that model's series for the hours of 11 November, with its hour. No average is taken and no model is preferred.

| Model (a model's output, read at Open-Meteo, evidence) | Run by | Run (UTC), as Open-Meteo reported it | Lisbon cell: highest gust value | Funchal cell: highest gust value |
|---|---|---|---|---|
| [ECMWF IFS](nodes/model-ecmwf-ifs.md) | ECMWF | [18:00, 10 Nov](https://api.open-meteo.com/data/ecmwf_ifs025/static/meta.json) | 108 km/h at 15:00 | 93.2 km/h at 05:00 |
| [GFS (NOAA/NCEP)](nodes/model-gfs.md) | NOAA | [18:00, 10 Nov](https://api.open-meteo.com/data/ncep_gfs013/static/meta.json) | 59 km/h at 13:00 | 86.8 km/h at 04:00 |
| [ICON (DWD)](nodes/model-icon.md) | Deutscher Wetterdienst | [18:00, 10 Nov](https://api.open-meteo.com/data/dwd_icon/static/meta.json) | 53.3 km/h at 16:00 | 159.5 km/h at 01:00 |
| [ARPEGE (Météo-France)](nodes/model-arpege.md) | Météo-France | [18:00, 10 Nov](https://api.open-meteo.com/data/meteofrance_arpege_world025/static/meta.json) | 81.4 km/h at 13:00 | 108 km/h at 04:00 |
| [Met Office Unified Model](nodes/model-ukmo-um.md) | Met Office | [18:00, 10 Nov](https://api.open-meteo.com/data/ukmo_global_deterministic_10km/static/meta.json) | 62.3 km/h at 12:00 | 105.5 km/h at 04:00 |
| [ECMWF AIFS](nodes/model-ecmwf-aifs.md) | ECMWF | not read | no gusts in its open output | no gusts in its open output |

A grid cell some 10 to 25 km wide is not a headland or a bridge, and a model's gust for a cell is not what IPMA's warning describes ("the most exposed places"). The models differ from each other. This site is not a warning service and does not forecast: for what to do, IPMA's warnings above and ANEPC.

## 3. The forecaster's product for pilots (TAF)

**Fact.** The aerodrome forecast for Lisbon airport, issued 11:00 Wed ([NOAA Aviation Weather Center](https://aviationweather.gov/api/data/taf?ids=LPPT&format=json), read 11:26 UTC, 11 Nov): strongest group **74 km/h gusting 102 km/h** (40G55 kt, probability 30%), 12:00 Wed to 16:00 Wed. 6 issues kept. A point forecast for one runway, written by a forecaster from the models above: quoted beside IPMA's warning, never instead of it.

`TAF LPPT 111100Z 1112/1218 14015KT 4000 RADZ BR SCT004 BKN008 TEMPO 1112/1116 14020G35KT 2000 RA BR SCT002 BKN004 FEW016CB PROB30 TEMPO 1112/1116 14040G55KT 1200 +TSRA SCT015CB BECMG 1116/1118 29012KT 9999 NSW SCT020 TEMPO 1118/1121 29017G27KT TEMPO 1200/1218 4000 SHRA BKN018 FEW020CB BECMG 1200/1202 23013KT`

## 4. The warning (IPMA decides)

**Fact.** [IPMA's weather warnings (avisos meteorológicos)](nodes/svc-ipma-warnings.md) for the Lisbon district ([IPMA](https://api.ipma.pt/open-data/forecast/warnings/warnings_www.json), read 11:26 UTC, 11 Nov) · [warnings.json](../api/v1/2026-11-10-storm-charles/weather/warnings.json). Highest level: **red**.

| Hazard | IPMA's level | From | To | IPMA's words | Sent by IPMA |
|---|---|---|---|---|---|
| rough sea | **orange** | 12:00 Wed | 21:00 Wed | Ondas de sudoeste com 5 a 7 metros e que poderão atingir até 13 metros de altura máxima. | 12:09 Tue |
| rough sea | **yellow** | 21:00 Wed | 03:00 Thu | Ondas de sudoeste com 4 a 5 metros. | 07:18 Tue |
| rain | **red** | 09:00 Wed | 18:00 Wed | Precipitação forte e persistente. | 21:23 Tue |
| wind | **orange** | 09:00 Wed | 21:00 Wed | Rajadas até 120 km/h. | 07:16 Wed |

Which model stands behind a warning? [IPMA's weather warnings (avisos meteorológicos)](nodes/svc-ipma-warnings.md): IPMA's page on its warning system gives thresholds, not models. A warning is a forecaster's decision over all of the above; that step is not published as data.

**Fact.** IPMA's communiqués kept: "Chuva, vento e agitação marítima - Depressão CHARLES - Continente e Madeira - Comunicado nº2" (issued 17:07 Tue); "Depressão BRUNA - Açores - COMUNICADO Nº3" (issued 08:30 Sat) ([IPMA communiqués](https://www.ipma.pt/resources.www/rss/comunicados.xml), read 11:26 UTC, 11 Nov) · [communiques.json](../api/v1/2026-11-10-storm-charles/weather/communiques.json).

## 5. The aggregator (MeteoAlarm)

**Fact.** The same warnings as CAP messages through [MeteoAlarm](nodes/svc-meteoalarm.md): 18 messages for Lisbon kept, sent between 12:58 Thu (2026-11-05) and 07:16 Wed (2026-11-11); 43 more for Madeira and other red areas ([MeteoAlarm feed for Portugal](https://feeds.meteoalarm.org/api/v1/warnings/feeds-portugal), read 11:26 UTC, 11 Nov) · [cap.json](../api/v1/2026-11-10-storm-charles/weather/cap.json). The CAP message carries the time IPMA sent it; IPMA's own open data does not, and lagged the CAP feed by two hours or more on 10 November (the data room's finding, in warnings.json → history).

## 6. What people looked at (sites and apps)

**Evidence (incident.weatherwatch@'s reading).** On 10 November, IPMA's own local forecast page, Yr and BBC Weather each printed an automatic wind for Lisbon far below the gusts of IPMA's red warning, with no explanation on the page: [research/weather-sites](../research/weather-sites/README.md). The chain explains why without anyone being wrong: those pages draw layer 2 (a model's mean wind for a grid cell), the warning is layer 4 (a forecaster's decision about the worst gusts in the most exposed places). Which model each site draws, where the site says: [IPMA](nodes/site-ipma.md), [Yr (MET Norway and NRK)](nodes/site-yr.md), [BBC Weather](nodes/site-bbc-weather.md), [Windy](nodes/site-windy.md), [Ventusky](nodes/site-ventusky.md), [MeteoPT](nodes/site-meteopt.md).

---
Not a warning service. For warnings follow [IPMA](https://www.ipma.pt/pt/otempo/prev-sam/) and ANEPC.