“Norton Summit” Details (Zwift Climb Portal)
Key Details
- Category: 2nd
- Length: 8.7 km (5.4 miles)
- Elevation Gain: 363 m (1,191‘)
- Average Gradient: 4%
- Country: Australia
- Watopia Climb Segment
- Watopia Descent Segment
- France Climb Segment
- France Descent Segment
- 5 W/kg: 19 minutes
- 4 W/kg: 23 minutes
- 3 W/kg: 28 minutes
- 2 W/kg: 39 minutes
Strava Segments
Time Estimates
The Tour Down Under describes Norton Summit as “Undoubtedly Adelaide’s favourite climb for cyclists.” This road is used as a benchmark climb for the city’s serious cyclists, but it’s also a gateway climb with many of the most popular riding areas in the Adelaide Hills accessible from the top.
This climb kicks off stage 3 of Tour Down Under 2025, described by race director Stuart O’Grady as “the toughest stage in the race’s history.”
We couldn’t find an IRL segment that matched Zwift’s start and end points, but the official Strava verified segment covers most of it (Zwift’s starts earlier and extends longer):
Why are gradients so different in both climbs? (set aside the start and finish points) Was because the GPS was doing funny things?
The climb seems great for a 20min FTP test
That’s a horrible attempt at matching the real climb. There’s no downhill sections after the switchbacks, and none of the very steep pitches. It looks like they’ve fallen off the side of the road and climbed back up.
Or is it the zwift version is reasonably accurate and the real world one has a bad elevation profile?
I rode it on another platform years ago and the ride really did fall off the edge of the road and climbed back up again. Don’t tell me Strava copied that profile! I’ll need to look properly when I’m not on my phone… There is no decent on this climb unless you include a final stretch to the post office, but most people consider the KOM at a crest before there.
This is that correct Strava profile (pretty much 5% the whole way): https://www.strava.com/segments/1013950
This segment stops at the ‘KOM’, which is roughly where the Veloviewer map shows the end (the post office is at the intersection at the end of the road). I hope the actual Zwift route is correct.
Reminds me quite a bit of the 6.9 miles climb out of Denbigh in North Wales, that ascends ~1200 feet overall and includes a one mile descent after the steepest section.
The aptly named Road to Hell
At first read I thought this was a PC management software product.
I find something from near Adelaide South Australia peaks my interest a bit.
There must be some good climbs in South America and Africa to get all the continents covered. I imagine Antarctica would be something of a challenge though…
I just climbed Norton’s other day while on holidays in Adelaide. This Zwift version isn’t the same climb that I did in real life.
in real life no downhill bits or real steep bits. Distance is about 5.5k
So Zwifts version is more accurate than the IRL elevation shown at the bottom of this article?
The zwift elevation profile is pretty close to the real thing but they’ve somehow made it 40% longer. The usual segment is 5.5km, the start and end points zwift have used would be 6.2km on the road, but they’ve somehow stretched that out to 8.7km
I’ve not ridden this, but the Zwift Climb Portal of Old La Honda Road in Woodside California, which I’ve ridden at least 100 times, also has problems. I’ve been actively engaged in route creation, extensively for the now dead RGT, and briefly for the new BikeTerra, and there’s ways to check profiles versus IRL: use data from riders with good bike computers from the actual segment as a check (not as a primary source). I just think Zwift doesn’t take the task seriously.
This page shows 2 different VeloViewer profiles. The top one is closer to the climb IRL. The bottom is what I climbed in Zwift. A local posted on Reddit that what is in the portal doesn’t look anything like the actual climb and after a little digging, it was revealed that there are 2 very different Norton Summit segments in Strava. Basically, the bad data was made without barometric altitude data. Bummer that the Climb Portal used the bad one!
Rims for road, discs for dirt, IYKYK. Smash the carbs. FB market place sl4 over Dentist sl8 any day of the week
Any idea why the Cauberg climb and decent is showing up on the Norton summit -for strava segments?
Not sure what you’re saying, but if you post a link to your Strava activity and explain what you’re seeing I can dig deeper…
As an example, here is my Strava activity on the Norton Summit with the Cauberg climb and decent showing: https://www.strava.com/activities/14542252120/
Woah, that’s interesting. It’s happening because the start of the Norton Summit Climb has basically the same GPS coordinates as the Cauberg! See Cauberg segment here: https://www.strava.com/segments/38937160
Rode this today (at 100%, rode down and exited the portal, before turning round and doing again at 75%) and no sign of a Strava segment registering and the leaderboard doesn’t want to register it either… Any ideas why?
Because if you ride it at less than 100% your Strava activity is marked as “gamified” and won’t be matched to any segments.
Oh drat… Which means I need to wait for it to come round in the rotation again, boo!
Wouldn’t it be great if we could ‘pick a portal’ on demand 😜