Eclipse 2026 Spain: Will You See It From Your Town?

Photo: Juan de Dios Santander Vela, CC BY-SA 2.0, via Wikimedia Commons.

On 12 August 2026 the Moon's shadow crosses northern Spain — the first total solar eclipse visible from the mainland since 1905, and the last until 2028. A Coruña, Oviedo, Santander, Bilbao, Burgos, Zaragoza, València and Palma all lie inside the path of totality — and the southern edge clips Madrid itself: eleven municipalities of the Comunidad are inside the band, though the city centre is not. Something like fifteen million people live within it.

Every guide tells you the same headline: totality falls between roughly 20:27 and 20:33 CEST, about an hour before sunset. What most of them mention only in passing is the thing that decides whether you actually see it. The eclipse happens with the Sun almost on the horizon — around 12° up in the far northwest, sinking to a couple of degrees by the time the shadow reaches the Balearics. At that height the sky is not the problem. The ground is.

The Spanish national astronomical body, the IGN, puts it plainly in its own advice: find a spot "with the western horizon, where the Sun will set, very clear of obstacles." That is easy to say and hard to check, because "clear of obstacles" depends on the exact shape of the land in one direction — the west-northwest — from wherever you happen to stand. So we computed it. For the entire path of totality, on 30-metre elevation data, we asked the only question that matters at this altitude: at the moment of totality, is the eclipsed Sun above the terrain, or behind it?

Why a low Sun turns geography into the whole game

When an eclipse happens high overhead, weather is the only variable. When it happens at 5°, a molehill matters.

The arithmetic is unforgiving. At 8° of solar altitude, an obstacle blocks the Sun once it rises about 140 metres for every kilometre between you and it. A 280 m hill 2 km to your west-northwest is enough. A ridge 400 m higher than you, 5 km away, is more than enough. And the further east you go along the Spanish path, the lower the Sun sits and the smaller the hill it takes.

Photo: Mentxuwiki, CC BY-SA 4.0, via Wikimedia Commons.

This is exactly the problem a terrain line-of-sight engine is built for, and it is the same one behind our study of the villages that lose the winter sun — only pointed at a single evening instead of a whole winter. For every point we build a 360° horizon profile from the elevation model, out to 50 km, with Earth's curvature and atmospheric refraction in the geometry. Then we drop the eclipse on top of it: using the real ephemeris we find, for that exact spot, the instant of mid-totality, the Sun's apparent altitude and bearing at that instant, and whether that point in the sky is above the local terrain horizon or below it.

We checked the eclipse geometry against the IGN's published local circumstances before trusting it on anything: for Oviedo the model puts mid-totality at 20:27 with the Sun 10.3° up, against the IGN's 20:27 and 10.2°; A Coruña, León, Zaragoza and Palma all land within the same tolerance. (Our totality durations run a few seconds longer than the official ones, because we treat the Moon as a perfect disc and skip the lunar-limb correction — it does not affect a single visibility verdict.)

The whole path, scored

Here is the result: every patch of land inside the path of totality, coloured by whether the totally eclipsed Sun clears its own terrain horizon or is hidden behind it.

Across the path, more than a quarter of the land — about 28%, some 60,000 km², an area larger than the whole of Aragón — cannot see the totally eclipsed Sun because terrain is in the way. Not cloud, not distance: rock and hillside, on any clear evening in history.

That figure is not spread evenly, and the reason is the single most useful thing in this article. In the west the Sun is still 10–14° up when totality strikes, high enough to clear almost everything; on the great tableland of Castilla y León it sails over an open horizon. In the east it is only 1–5° up, and the landscape reaches it. The blocked fraction climbs almost monotonically from one end of the country to the other:

So the good news and the bad news are the same fact. If you are in the western half of the path — Galicia, Asturias, León, Castilla — the odds are strongly with you, and this article is mostly reassurance. If you are heading for the Mediterranean end — Castelló, València, the Balearics — the odds are that some specific spot near you is ruined, and picking the right one is worth minutes of planning. The rest of this piece is the specifics.

The same gradient shows up city by city, and it is the whole reason the east has to plan while the west can relax:

Play with the whole path

The map above is a snapshot; this one you can drive. Switch between the terrain verdict and the Sun's altitude, drag the slider to demand more clearance and watch the safe ground shrink, and click anywhere for that spot's numbers and a link to run it yourself. It now reaches half a band width beyond each edge of the path as well, in paler colour: out there the eclipse is not total, but the Sun is still 98% covered or more and the same ridges still take it.

This map also lives on its own, full screen, as the Eclipse Horizon Checker. Same data, more room.

Why the last 1% is not a rounding error

The map prints numbers like 99.8% covered, and it is worth being blunt about what that means: it is not 99.8% of a total eclipse. It is not close.

The Sun's surface does not dim as the Moon slides across it. Every scrap of photosphere still showing is exactly as bright as it was at noon, so what you lose is area, not intensity. Full daylight is around 100,000 lux. Cover 99% of the disc and roughly 1,000 lux is left, which is a bright room. Cover 99.9% and you are near 100 lux. Totality is about 1 lux, in the region of moonlight. Between the ring on this map and the band inside it there is a factor of a thousand, and factors of a thousand are not subtle.

That factor is the whole experience. The corona only appears once the photosphere is completely gone; at 99.9% the remaining thread outshines it entirely and you will never know it was there. The horizon does not turn into a 360° sunset, the temperature drop is muted, the planets do not come out, and the shadow does not visibly race across the landscape at you. You also cannot take the filter off: a sliver of Sun is still a sliver of Sun, and it will hurt your eyes at 99.99% exactly as it would at noon.

Inside the band, all of it switches on at once, for as long as the duration in the popup says. That is why the paler ground on the map is drawn paler, and why being 20 km on the wrong side of the limit is worth an hour in the car.

The meseta wins

The flat interior is the easy winner, and it is worth saying plainly because it runs against instinct: the boring landscape is the best one. On the wheat plains around Valladolid, Palencia and León the western horizon is a straight line, the Sun is still 8–9° up, and totality is simply there for the taking. Soria clears its horizon by more than five degrees; Zaragoza, from beside the Basílica del Pilar, by six; Guadalajara, Cuenca and Lugo by five to nine. These are not close calls. They are open sky.

Photo: David González Gómez, CC BY-SA 4.0, via Wikimedia Commons.

Even here, though, the computation catches something a map would not. Take Burgos. Stand by the cathedral and you catch totality by a whisker — the eclipsed Sun clears the skyline by under a degree. Walk a few streets west, down into the shadow of the Cerro del Castillo, the wooded castle hill that rises straight off the old town, and it is gone: from the low ground the hill stands seven degrees up in exactly the wrong direction. Climb onto the castle hill instead and you clear the horizon by eight degrees, with the whole meseta laid out beyond. One city, three verdicts, inside a fifteen-minute walk. This is the recurring lesson of the whole study — a town is not a point — and it is why the table at the end lets you check a specific place rather than trust a region.

Photo: BurgosBus, CC BY-SA 3.0, via Wikimedia Commons.

The mountains take their share

Where the path crosses real relief, real totality goes missing. The band clips the Cantabrian Cordillera along the north coast, runs the length of the Sistema Ibérico through Soria and Teruel, and finishes in the coastal sierras behind Castelló and València — and every one of those ranges paints a swath of the map red.

The sharpest single example is in the Picos de Europa. Potes, the little capital of the Liébana valley, sits in a bowl under the Andara massif, and on 12 August it lands on the knife-edge: the eclipsed Sun clears its skyline by six-hundredths of a degree, then slips behind the wall while totality is still running. The Sun is up — 9° above the flat horizon — but the mountain is 9° up too, in precisely the direction the Sun has drifted to.

Nearby, the story is the same wherever the walls close in. The top of the Fuente Dé cable car — 1,800 m up and feeling like the roof of the world — is actually hemmed in: the cirque wall to its west stands nearly two degrees above the eclipsed Sun. To keep totality in the Picos you have to get above the ridgelines, not just high: the summit of Torre de Cerredo clears everything by ten degrees, and the Covadonga lakes, open to the northwest, clear by four. High is not the same as open — the mountain has to be below your sightline, not merely beneath your feet.

Photo: Master08, CC BY 3.0, via Wikimedia Commons.

The coast: go to the beach

On the Mediterranean end of the path the Sun is desperately low — 2 to 5° — and the coastal sierras stand right behind the towns. This is where terrain does its worst, and where the fix is simplest: get to the open water.

The pattern repeats down the whole coast. València's old centre loses the eclipse; its Malvarrosa beach, looking clean out to sea, keeps it by more than three degrees. Castelló the same: blocked in town, wide open on the Grao seafront. Benicàssim is genuinely stuck — the Desert de les Palmes sierra sits six kilometres to its west-northwest and takes the Sun a full degree early. Up the coast, even Santander loses totality at the point our gazetteer calls the city centre, by a hair — yet a short walk to the Sardinero beach or out to Cabo Mayor lighthouse turns a marginal miss into a nine-degree certainty. On this coast the rule is one line long: stand where you can see the sea to the west, and let the water be your horizon.

Palma: the line runs through the city's own beach

The single most consequential result in the study is Palma de Mallorca, because it is the largest city in the whole path — 438,000 people — and it is on the wrong side of the margin.

At Palma the eclipsed Sun is only 2.6° up, and the hills west of the city — the Na Burguesa ridge that Bellver Castle sits on, and the Serra de Tramuntana rising behind — are just tall enough to catch it. From the cathedral and the Passeig Marítim, from the whole western bay through Palmanova and Magaluf, totality is lost behind that terrain, by a few tenths of a degree to a couple of degrees.

But the map holds a reprieve, and it is a strangely intimate one. The line between losing and keeping totality runs straight through the city's own seafront. At the cathedral you miss it by half a degree. Two and a half kilometres east at Portixol you scrape it. By Can Pastilla and out along the Platja de Palma toward s'Arenal — the long resort beach most of Palma's visitors are already standing on — the eclipsed Sun clears the sea by a full degree and more. Half a million people, and the difference between darkness and disappointment is which end of their own beach they walk to.

Photo: dronepicr, CC BY 2.0, via Wikimedia Commons.

The other islands split the same way. On Ibiza, the town itself is blocked by a low hill to its west, but Sant Antoni bay, Cala Comte and the Es Vedrà miradors — the whole west coast, which is where you would go to watch a sunset anyway — are wide open over the sea. Formentera is clear. On Menorca the open western and southern shores keep it — Ciutadella clears by well over a degree — while Maó, lower and hemmed in to the east, loses it; totality is short across the island either way. On every island the instinct that sends you to the western shore for the sunset sends you to the right place for the eclipse.

The line of totality runs through Madrid

This is the part almost every eclipse guide gets wrong, and we got it wrong too in an earlier draft. Madrid is usually written off as "outside the path". The city centre is: from Sol or Atocha you get a deep partial and nothing more. But the southern limit of totality does not stop politely at the province border — it cuts across the north of the Comunidad, and then across the city itself.

Eleven municipalities of the Comunidad de Madrid sit inside the band. Here they are with the duration of totality we compute, and — the part nobody else checks — whether the terrain actually lets you see it:

Look at La Cabrera. It is on the official regional list of towns that get totality, it gets a comfortable 78 seconds of it — and from the town centre a ridge is in the way by a tenth of a degree. Being inside the path is a necessary condition, not a sufficient one. That is this whole article in one row.

Inside the city of Madrid the line is even finer. North of roughly the M-30 the band clips the top of the municipality: Barajas grazes about 36 seconds, Fuencarral–El Pardo about 27, Chamartín about 24, Ciudad Lineal about 22. Sol, Atocha and everything south of them get nothing.

Two honest caveats, because this is the graze zone and precision matters most exactly where it is hardest. First, at the very edge of the band totality duration is extremely sensitive to the model: ours run a few seconds long because we do not correct for the Moon's ragged limb, and published figures for the same town disagree with each other by half a minute. Second, the Comunidad's official list stops at municipal boundaries and does not include Madrid capital. So treat the city districts as seconds, if the limb cooperates — not as a plan. If you want the real thing rather than a rumour of it, drive 40 minutes north.

Three rules, from 300 million measurements

Everything above is one place at a time. Cross the whole raster against the shape of the ground and three rules fall out — and none of them is "go high".

Flat beats high. Ground with less than 2° of slope loses totality only 4.8% of the time. A 20–35° hillside loses it 60.2% of the time, twelve times more often. And height itself does not save you: the safest band is 800–1000 m, where 21% is blocked, while above 1400 m it climbs back to 46%. Being high is worthless if the things around you are higher. Fuente Dé's cable-car station sits at 1,850 m and still loses the eclipse to its own cirque.

Face west, not east. This is the strongest single signal in the dataset. On a slope that faces west — you are looking down the hill in the direction of the setting Sun — only 12.1% of the land is blocked. On a slope that faces east, with the hill rising behind you exactly where the Sun will be, 64.1% is blocked. Same mountains, same eclipse, five times the risk.

If you are blocked, walk. This is the hopeful one. We ran a distance transform over all 82 million blocked cells: 56% of blocked ground has clear ground within 100 metres, and 88% within 400 metres — five minutes on foot. Losing the eclipse to a hill is almost never a property of your town. It is a property of the exact spot you happened to stand on, and the fix is usually to walk to the end of the street.

The toll, counted

Add it up across the whole path and the shape is clear. Of the 4,100-odd towns inside totality, most are fine — but for 104 towns, home to about 874,000 people, a mountain two or more kilometres away deletes the totally eclipsed Sun from the centre of town. Palma dwarfs the list; behind it come a string of Balearic and coastal towns, and a scatter of valley villages in the Pyrenees, the Iberian System and the Picos where a single wall does the whole job. Hundreds more towns lose it to nearer, smaller rises — the kind of local relief that is real in a steep village and, in a flat city, is as likely to be a building as a hill, which is why we count the distant mountains separately and send you to a real vantage point for anywhere urban.

None of this is a reason to stay home. It is a reason to move a few kilometres before you commit — and, for a once-in-a-generation event with hard-booked accommodation, to do it before you book, not on the day.

Check your own spot

The map above is the whole country at a glance. The table below is the specific answer for every town of 3,000 or more people inside the path — 554 of them — with the Sun's altitude at totality and whether the eclipsed Sun clears the terrain at the town centre. Sort it, filter it for your town, and open any row straight in the calculator with the pin already dropped.

Every town of 3,000+ people on the map — 1,260 of them, inside the path and for half a band width beyond each edge — with how much of the Sun goes and whether the eclipsed Sun clears its own terrain at the town centre. Sort it, filter it, or open any town on the checker map with its pin already dropped. “Sun covered” is the fraction of the solar disc the Moon takes at maximum eclipse: 100% is totality, and 99.8% is not. “Margin” is how far the eclipsed Sun sits above (＋) or below (−) the terrain horizon; the town centre is the GeoNames coordinate.