Blog

How Accurate Is a GPS Watch Under Tree Cover? | Suunto Vertical 2, Measured on Mount Nikko-Shirane (2,578m)

2026.08.24

The scariest thing in the mountains isn't getting lost, exactly. It's not knowing where you are right now.

You stand at a junction, you read the sign, and you still aren't sure. Those thirty seconds are the loneliest part of any hike. I suspect I'm not the only one.

When I used the Suunto Vertical 2 on Mount Amagi last time, the thing I kept wondering about was GPS accuracy. "It seems about right" is a weak thing to say in a review. So this time I walked Mount Nikko-Shirane (2,578m), the highest peak north of Tokyo, from morning to evening — and checked it with numbers.

The Suunto Vertical 2 was provided by Suunto. Every figure below comes from that day's log, exported from the Suunto app. I haven't cherry-picked the flattering ones.

Why measure under tree cover

The most important point first: if you test GPS on an open ridge, the answer will be "it barely drifts" — and you'll have learned nothing.

GPS works by reading signals from satellites overhead. Trees and rock faces block those signals or bounce them around, and that's where positions start to wander. The value of dual-band GNSS only shows up in bad conditions.

Nikko-Shirane is a good test site for exactly that. From the ropeway station at 2,000m you climb into forest where the sky is almost entirely hidden. Above the treeline it flips completely — bare rock, nothing overhead. The worst conditions and the best conditions, in the same day.

Three things to do before you leave home

The preparation is genuinely just this:

1. Download the topographic map for that mountain to the watch over home Wi-Fi 2. Draw your route and send it to the watch 3. Check the day's climb with Climb guidance (the route profile)

All three take a few minutes at home. Try to do them at the trailhead and you'll find there's no signal, so this is the one part worth finishing the night before.

How far off was the position at the junctions?

We started walking at 9:08 and were back at 15:39. Six hours 31 minutes, 7.47km, 616m of ascent. The watch ran all day in Performance (multiband) — the mode that eats the most battery, and the one I wanted to test.

I then overlaid the recorded track on the route I had planned in advance, and calculated how far every recorded point sat from that planned line.

| Deviation from the planned route | Measured | |---|---| | Median | 2.6m | | Within 10m | 97.5% | | Within 25m | 100% | | Maximum | 19.2m |

And at the junctions inside the forest — the numbers I actually cared about:

• Junction 1 (deep in the trees): 3.7m
• Junction 2 (at the trail sign): 7.0m
• Summit: 1.0m

A few metres is about the width of the trail itself. For looking at the map screen and deciding "yes, I'm on this path," that is more than enough.

The watch's own accuracy estimate (EHPE) agrees: a median of 3m, under 5m for 90% of the day, 13m at worst. It was holding a median of 31 satellites (17 at the fewest, 36 at the most) — under a canopy, all day.

I climbed without calibrating the altimeter

The second test was altitude. Normally you'd tell the watch "we're at 2,000m" before setting off, and everything after that reads more accurately. I deliberately skipped that step.

The marker at the trailhead says 2,000m. The watch said 1,998m — a 2m gap from the start. I left it alone and walked for six and a half hours.

• Start (ropeway station, marker at 2,000m) → watch: 1,998m (−2m)
• Summit (true elevation 2,578m) → watch: 2,583m (+5m)
• Back down → +3m

One caveat worth understanding: this watch doesn't derive altitude from the barometer alone. It uses FusedAlti, which blends GPS altitude with the barometric reading. So the usual framing — "how far had it drifted X hours after calibration?" — doesn't really apply here. Read the numbers above as the difference at each moment, rather than as accumulated error.

Either way: no calibration, a full day of walking, and 5m at the summit. That's plenty for how I hike.

Battery: 99% → 75% in six and a half hours

A full day in the most power-hungry mode, opening the map and the navigation screen constantly. The log says:

99% at 9:08 → 75% at 15:39 — 24 points.

Extrapolated, that's roughly 27 hours of the same kind of use. I never once thought about the battery during the day, and I'd happily do an overnight trip without carrying a charger.

(The catalogue figure is up to 65 hours, but that's measured under different conditions. Treat the number above as "a day hike where I used the map and navigation as much as I wanted.")

Who this watch is for

After a full day with it, these are the hikers I think it genuinely helps:

• People who hike forested mountains, or trails where the route isn't obvious. Conditions like these are exactly where the difference shows
• People who already plan their routes in advance. The more of that three-step prep you do, the less there is to think about on the mountain
• People doing day hikes and overnights who don't want to carry a charger. At this drain rate, you simply don't think about it

And to be fair, some people don't need to add one:

• If pulling the phone out of your pack has never really bothered you
• If you mostly walk well-maintained low mountains where getting lost isn't a worry

You don't need a watch to climb a mountain. But the time you used to spend opening your pack to check where you are does disappear — and that changes how the day feels. Whether that's worth it to you is the real question.

The part I want to be honest about

Those are good numbers. I still don't think you should lean on GPS alone.

Batteries die. Watches get dropped. In conditions like these, there will be moments when the numbers wobble. Carry a paper map, a compass, and a phone with a mapping app — always. On top of that foundation, the watch earns its place by removing friction, not by replacing judgement.

And one more thing: this watch does not make navigation errors disappear. Knowing where you are and making the right call are two different skills. Even on this hike, the sequence was: stop at the junction, read the sign, *then* check the wrist.

Doing the same thing since 1936

At the summit there was a moment with a compass made ninety years ago. Suunto started in Finland in 1936, making liquid-filled compasses.

I don't think the job has changed. Build a tool that tells you where you are, in the mountains. The insides went from a magnetised needle to dual-band GNSS, but the question it answers is the same one. Thinking about that on the summit made the watch on my wrist look a little different.

In short

How accurate is GPS under tree cover? On this hike: a median deviation of 2.6m from the planned route, 97.5% of it within 10m, 3.7m and 7.0m at the two forest junctions, 1.0m at the summit. Altitude, uncalibrated, was +5m at the top. Battery cost 24 points over six and a half hours.

But the number I'd actually sell you on isn't in that list. It's how much less time you spend standing still at junctions. No pulling the phone out of the pack — just a turn of the wrist. Repeated a dozen times a day, that adds up far more than I expected.

I've written about the same watch in the fog on Mount Amagi, under completely different conditions. Worth reading alongside this one.

Related posts

Do You Really Need a GPS Watch for Hiking? | 5 Features, Tested on Mount Amagi

8 Pieces of Gear I'm Glad I Bought in My First Year of Hiking | An Honest Best-Buy List

A Whole Summer in 10 Montbell Pieces | An Honest Minimal Wardrobe

A Complete Guide to the Koboyama Hike | Hadano Station to Tsurumaki-Onsen Station, Three Gentle Hours with Binoculars

This post contains affiliate links. Prices and specifications are as of the video's publication date. Most items are sold in Japan.

← All posts