Skip to content

Tool

Calima: when the Saharan dust actually arrives

Calima is the one Tenerife weather event that can actually spoil a week, and it is the one every guide describes with a shrug. Here is the frequency, the intensity and the duration, by month and by zone.

Pick a zone

Dust and PM10 by month for the selected zone
MonthDays over EU PM10 limitNoticeable dustHeavy dustMedian PM10Worst 10% of daysDays
January19%26%10%18 µg/m³120 µg/m³93
February20%27%12%18 µg/m³130 µg/m³85
March1%1%0%11 µg/m³19 µg/m³93
April8%10%0%13 µg/m³37 µg/m³90
May0%0%0%11 µg/m³18 µg/m³93
June0%1%0%12 µg/m³22 µg/m³90
July2%2%0%13 µg/m³27 µg/m³93
August12%18%0%17 µg/m³56 µg/m³93
September7%8%1%12 µg/m³31 µg/m³90
October9%14%3%13 µg/m³38 µg/m³93
November14%19%2%13 µg/m³88 µg/m³90
December30%40%11%17 µg/m³126 µg/m³93

“Noticeable dust” = daily mean Saharan dust over 50 µg/m³; “heavy” = over 200 µg/m³. The PM10 column uses the EU daily limit value of 50 µg/m³, which is a health standard, not a visibility one — a day can cross it and still look merely hazy.

How long an episode lasts

Frequency alone is misleading: what ruins a trip is not a dusty afternoon, it is a dusty week. These are runs of consecutive days above the EU limit.

ZoneEpisodes per yearTypical lengthLongest observed
South resorts (Costa Adeje / Los Cristianos)122 days18 days
North coast (Puerto de la Cruz)9.72 days17 days
Santa Cruz / La Laguna9.72 days17 days
Las Canadas / Teide (2.000 m+)9.72 days17 days

What the record says

The result that surprised me is the seasonality. The story you hear repeated is that calima is a summer thing, arriving with the heat. In this record the worst month on the south coast is December, with the EU daily PM10 limit exceeded on30% of days, followed by February at 20%. August is real but it is not the peak.

This is not a contradiction of the summer story so much as two different phenomena wearing the same name. Summer calima tends to ride in high and hot, which is why it is memorable; winter dust arrives in a lower, cooler layer that hangs around the coast. The health metric used here — daily mean PM10 at ground level — picks up the second one more readily than the first.

At the other end, 2 months (May, June) recorded no days at all above the limit in this sample. With only 3 years behind it, read that as “rare”, not “impossible”.

On duration: the south coast sees about 12 episodes a year, the typical one lasting 2 days. But the tail is long — the worst run in the record lasted 18 days. That asymmetry is the practical point: the median episode is a nuisance you can wait out, and the tail is the one that eats a holiday.

The honest limitation

Three years is a thin sample for monthly statistics, and I would rather say so than dress it up. The ranking of months is probably stable; the exact percentages are not. Extending this back to 2003 is possible using the Copernicus CAMS reanalysis directly rather than through a convenience API, and that is the next thing worth doing to this page.