Seville Heatwave: AEMET Forecast Overshot Reality by Six Degrees

The heatwave that, according to a forum user, AEMET got wrong four out of seven days

Monday, August 4, forecast 39 degrees, actual reading 39.8. Tuesday, exactly 39. Wednesday, 37.6 against the announced 39. The start of the major heatwave expected to scorch Spain matched almost to the decimal. The problem arrived on Friday the 8th: the forecast indicated 41 degrees but the mercury stayed at 35.3. A six-degree difference. Saturday saw 42 announced and 38.8 real. Sunday had 44 predicted and 41.3. Four out of seven days with an upward deviation, all in the same direction.

The question hovering over this assessment was simple: what model does AEMET use that always errs upwards? Because a systematic bias, argues the harshest analysis, is not a technical error, it is a decision. And if the model had such a bias, logic dictates correcting it.

What exactly failed in the heatwave prediction

The count is telling. Three days accurate or nearly so (Monday, Tuesday, Thursday), one moderate miss on Wednesday, and three severe upward deviations from Friday to Sunday. The sum paints an uncomfortable pattern: when the model makes mistakes, it heats up the atmosphere. There is not a single day where the prediction fell below reality.

Some nuance that forecasts are issued for specific locations and topography introduces variations of several degrees between towns five kilometers apart. This is true and is the most reasonable argument. It is also true that a six-degree deviation is not explained by a valley.

The heatwave narrative versus thermometers

The underlying suspicion is not meteorological, but communicative. For a week, there was relentless focus on the heatwave, warnings, and hellish headlines, only to find the unbearable heat arrived late. The first day Seville exceeded 40 degrees was precisely the one closing that warning cycle. Again.

The contrast with the rest of summer fuels the thesis. In July, the same channels repeated that temperatures were below normal, with days up to ten degrees under average. Now, any tenth becomes a historical record. The asymmetry — cool when it should be cool, apocalypse when it should be apocalypse — is what causes the most discomfort.

Seville data debunking the record

The historical review of Seville airport deflates the epic. In 1994, 40 degrees were exceeded on 16 days, with a high of 43.9. In 1995, 15 days and a peak of 46.6. In 2022, 20 days and 44.8. In 2023, 18 days and 43.7. In 2024, 8 days and 42.2. This year, between July and August, there have been 9 days above 40, with a maximum of 44.2.

The conclusion is uncomfortable for both sides: the heat exists, is intense, and fits within the historical series. Seville's absolute record has stood unbroken for 30 years. Even the 2021 record of 44.9 has not been beaten. The most advertised heatwave was not the hottest.

Air conditioning and heat memory

There is a less conspiratorial explanation: we have become acclimatized poorly. In 1980, people endured heat from sunrise to sunset, without air conditioning at home, in cars, offices, or shops. Today we move from bubble to bubble, and ten minutes outside feel unbearable. Perception rises even if the thermometer doesn't.

Another part of the analysis looks at underlying data. In the Ebro Valley, June, July, and August series from the seventies, eighties, nineties, and two-thousands show a clear rise in highs, lows, and averages. In Madrid, the normal range in childhood was 33-35 degrees; today 36-38 is standard. In La Mancha, the threshold for 'very hot' moved from 35 to 40. In the Upper Ebro, however, Vitoria's minimums still drop below 15 degrees and tropical nights are rare.

---

The episode's close left a less epic headline than expected: models were right for the three days of true intensity, the maximum stayed at 44.2, and Seville's absolute record remains intact after three decades. The heatwave existed. The record did not.