Ken Roberts - - Bicycling

what's here

later in year

see also:  more Europe stories | public discussion | more on Europe

trickiness measuring steepness for Monte Zoncolan


A month later I found yet another elevation profile.


After writing the analysis below, I found this:

elevation profile published by Stefano Orazzini (similar profile on

Here's a list of other profiles on that site.


The west side road on Monte Zoncolan near Tolmezzo in Friuli in northeast Italy [ see where on map ] is a notorious big steep road climb - (by some criteria, the steepest big asphalt-surface road climb in Europe or North America). I decided that I would be unable to pedal my bicycle up that road without needing to get off and walk in two sections -- because I analyzed a steepness profile graph on for that road which has been linked or its steepness numbers used widely around the web (including Italian and English language versions of the Wikipedia).

But then I made to the top pedaling with no problem -- which made me suspect that some of the steepness numbers were not correct. Which fit with the difficulties I've had using different methods to measure the steepness of two hills closer to home.

I think these same kinds of difficulties apply to measuring the steepness of lots of other roads with are curvy and steep.

my puzzle

That elevation profile graph showed one section of about 0.75 km of horizontal distance at steepness over 19% grade (i.e. 133 vertical meters of climbing in 700 meters of horizontal distance), and another section of about 0.38 km of distance at steepness over 20% grade (i.e. 76 vertical meters of climbing in 375 meters of horizontal distance). (Italian and English language Wikipedia articles as of September 2008 say 20%). The problem I felt is that the 19% section in that profile is more than twice as big as the longest 19% section I've ever pedaled my bicycle up before in my life (on a much smaller overall hill). So it seemed unlikely that could make it up the Monte Zoncolan on my bicycle the whole way pedaling.

Actually in September 2008 on my first time trying Monte Zoncolan, I made all the to the top pedaling -- and I didn't notice any section anything like 133 vertical meters  that seemed much steeper than most of the rest of that road. At my climbing speed, 133 vertical meters should take me at least 8 minutes to pedal through, so it's not like I might have just missed it. And on other climbs I very easily notice the difference between 14% and 19% grade sections -- which on the elevation graph is the transition shown from the section before each steep section.

Also it seemed to me like the road on the west side was overall carefully engineered to have a fairly consistent steepness. Constructing a road like that takes months, so the engineers have plenty of time to carefully survey the intended route with accurate instruments. Some hills have difficult geological features like rock outcroppings and cliffs that make it difficult to satisfy an intended design constraint, but it didn't seem to me that the west side of Monte Zoncolan had those problems. To me it seems unlikely that otherwise careful road engineers would just sorta "forget" their goal in two major sections.

trickiness of measuring steepness

Steepness of a road is tricky to get accurately, because you have to measure two things: altitude and distance (not in a staight line) -- and you have to coordinate the two kinds of measurement.

Really the problem is that measuring even one kind of quantity in the real world is actually trickier than most people think. There's two main problems:

(a) most people don't even try to follow procedures in measuring which could help deal with the trickiness;

(b) successfully following good measuring procedures requires at least two of these: time, care, money.

Applying those to curvy steep roads:

(a) most bicyclists don't even know that there's a question of trickiness -- just show them a graph from topo software or a GPS or some web page, and they just accept it.

(b) getting more accurate measurements of steepness for bicycling purposes is just not worth putting in the time + care + money to achieve.

I agree that getting the accuracy not worth it for bicycling.


  • I am not surprised when I see two elevation profiles for the same climb which obviously disagree.

  • Elevation profiles from topo software or websites are OK (or sometimes not OK, depending on the source) for getting some sort of idea about if a road is not steep or sorta steep or very steep. But the simplest answer is . . .

  • just go to that road and ride up it.

(for more accuracy, ride up it several times on different days)

Monte Zoncolan:

  • Just because some detailed full-color graph shows that it has two major sections at 19-20% doesn't mean they're really there out on the road.

  • Just because the 20% grade number has been uncritically copied into several magazine articles and Wikipedia articles about Monte Zoncolan doesn't mean there's a section that steep while riding up the actual road.

  • I doubt there's any sustained section on it which is steeper than 16% grade.

(though by "fractal geometry" it is surely possible to find a very short section as much steeper than that as you desire -- but that has no relevance to bicycling)

qualifying that Mt Z elevation profile data set

I'm really glad and grateful that Gianpaolo Vicario gathered that sequence of altitude and distance data for the west side of Monte Zoncolan and made it public on this web page.  It was way better than anything which came before. I think it was appropriate for him to present the data just the way he did.

What is not appropriate is that the Wikipedia uncritically used some results from it without qualification. But that's not Gianpaolo Vicario's fault. And how could he have guessed that everybody else who came after him would be too lazy to measure that road again carefully themselves?

Here's the problems (or qualifications) of that data set not mentioned by Wikipedia (and others) who used results from it:

  • the measurement instruments or sensors for distance were not disclosed.

  • the measurement instruments or sensors for altitude were not disclosed.

  • the overall procedure of measurement was not disclosed.

  • no mention of a second person present to help "vet" the procedure and keep it on track.

  • digital averaging + "cleaning" algorithms built into the sensors were not discussed.

  • the date(s) was not disclosed.

  • the accuracy of instruments and procedures for measuring distance was not characterized (i.e. the statistics of variance and skewness of measuring in a situation whose true distance is known more accurately by other means).

  • the repeatability of the instruments and procedures for measuring distance was not tested or characterized.

  • the accuracy of instruments and procedures for measuring altitude was not characterized (i.e. the statistics of variance and skewness of measuring in a situation whose true distance is known more accurately by other means).

  • the repeatability of the instruments and procedures for measuring altitude was not tested or characterized.

  • no mention of attempting to repeat the overall measurement process a second time on the same day.

  • no mention of attempting to repeat the measurement process on a different day.

  • no mention of attempting to repeat the measurements with a different kind of instrument (or at least a different manufacturer's brand).

Again, this is not a criticism of the original work of Gianpaolo Vicario, for which I am only grateful.

Rather this is a criticism of Wikipedia articles about Monte Zoncolan (English and Italian language versions of Wikipedia as of September 2008).

Those Wikipedia articles:

  • did not attempt to follow or investigate following procedures which are standard in some other parts of the Wikipedia for quoting measurements.

  • did not attempt to qualify the measurement claims or sources.

  • did not attempt to compare the measurement claims against other independent sources -- some of which require less work and cost (e.g. topo software, Google Earth).

problems with using topo software + Google Earth to measure steepness

Once you learn how to use the software, this is the easiest of all -- you don't even have to go out to road with the hill.

But it has serious problems -- so it's nice for generating ideas for climbs to try -- but always needs to be checked against data from other sources -- and usually modified to take into account what's learned from those sources.

Problems to watch out for:

  • altitude data for topo software and Google Earth tends to be smoother than the real world for complicated terrain like that often found around steep curvy roads.

(other things being equal)  This tends to make a road get calculated as less steep on topo software than it is in the real world.

  • precise locations of roads can be difficult to get accurately -- especially for smaller and narrower roads which are often interesting for bicyclists. Creators of mapping data typically get road location data from aerial photographs. But smaller and narrower roads are often obscured by trees in aerial photographs -- so the mapmakers (or their software) have to guess about the precise location of the center of the road. But on a steep hillside a small difference in the location of the center of a road can result in a large difference it its calculated altitude - (even worse when the road is cut into the side of a cliff).

Sometimes on Google Earth when I overlay the curved line for a small road onto the aerial photograph, I can see that the road line doesn't fit what's plainly visible in the photo -- but there could be other complications in that (e.g. the aerial position and relative angle from which the photo was shot).

(other things being equal)  This tends to make a road get calculated to have more variations in steepness on topo software than it has in the real world -- which implies that a climb on topo software will tend to show steeper short sections than it has in the real world.

Of course the mapmaker could get more accurate location points by just sending somebody out to drive the road with a GPS -- but that gets very expensive when there's millions of kilometers of narrow roads to be mapped.

  • road paths for topo software tends to be either a bit smoother or a bit straighter than roads in the real world. Depends on how the software stores its data: as line segments or curves. Line data is straighter -- until it makes a discontinuous bend to the next segment.

Engineers who are designing and constructing an actual steep road tend to make small variations in its path to work around small but difficult terrain features - (unless they're being forced to comply with some line drawn on a map by some bureaucrat sitting in an office in a distant city who doesn't know or care about the real world of a little road.)

Topo software is more like a distant bureaucrat drawing a pretty curve on his map.

Since competent road engineers tend to try to construct roads with more consistent or smoother steepness - (but also complying with other constraints like cost and time) . . .

(other things being equal) -- Topo software which tends to ignoring those clever little variations will tend to make a road get calculated to have more variations in steepness than it was actually constructed by a competent engineer out in the real world -- which implies that a climb on topo software will tend to show steeper short sections than it has in the real world.

  • the algorithms used to detect or store or generate road paths may have deliberate distortions which impact steepness calculations.

Like some mapping software versions (not necessarily including topo data) I know systematically overstate distances of roads -- perhaps because very few customers will complain that a journey took a little less time than they expected based on the software.

I could believe (but have not verified) that some topo software algorithms systematically overstate the total climbing of a road that has ups and downs -- perhaps because very few customers will complain that traveling over a road took a little less effort or time than they expected based on the software.

One fundamental factor about topo software which makes roads seem smoother with fewer steep sections, and two factors tend to make roads seem to have more varied grade, with some sections steeper than in reality. Overall I think for steep curvy roads, topo software tends to show more steeper sections than are found in the (humanly engineered) road out in the real world.

Fundamental problem with relying on topo software for calculating steepness of a road sections: You know there are distortions, but you don't what they are or in which direction -- because you haven't been actually gotten out there on the real road.

problems with using GPS to measure steepness

Using a GPS is basically easy -- just take it along with you for the climb, start recording at the bottom, stop recording at the top, afterward upload it to your computer or some website.

The basic problem with most GPS units is that the whole system of GPS satellites and their data is simply not set up to accurately measure altitude -- but that's one of the two fundamental components for calculating steepness.

Therefore altitude date and steepness profile from any GPS which does not have a barometric altimeter is suspect.

Most GSP units do not have a barometric altimeter. A good barometric altimeter tends to be expensive. Therefore . . .

  • If you're not using an expensive GPS - (and one which specifically says that it has a barometric altimeter) - you're likely getting significant inaccuracies in the steepness profile of a steep curvy road.

  • A GPS without a barometric altimeter is likely OK for getting the overall average steepness for a long climb -- but don't believe what it says about variations among shorter sections.

Other problems to watch out for:

  • modern GPS units use various numerical "algorithms" for averaging and smoothing and "cleaning" the raw data from the satellites -- and an algorithm for converting a series of distinct location points into a distance measured along a curved path. On a steep road with variations and curves, these algorithms could be a problem, because (a) the manufacturer has likely "tuned" the algorithms for smoother terrain and straighter roads and higher speeds, and because (b) the algorithms are unlikely to have been tuned for the purpose of measuring steepness. Then since you likely don't know what those smoothing + "cleaning" algorithms are, you don't know how to compensate for them.

  • If the GPS directly displays some number for steepness, I would likely not rely on that. Since I don't know how it's calculated - (and it might even be deliberately distorted by the manufacturer).

  • If you weave from side to side while you ride up the hill, the distance recorded by GPS might be longer -- or it might not -- depends on what its algorithms are. If the distance recorded by the GPS is longer than if you had ridden straighter, then the steepness you calculate afterward will be lower. Which is more accurate for your actual riding, but less accurate for measuring the steepness of the overall road.

  • Lots of interesting steep curvy roads have big trees alongside and hanging over them. Tree leaves might interfere with GPS satellite signals. For better accuracy on such roads, might try measuring in the winter.

  • if a weather front or storm is approaching, the barometric altimeter might not be accurate for a long climbing section. (but I doubt this is usually a major problem for measuring steepness of each specific shorter section of a climb). It's a good idea to have a measurement of the difference in altitude between the top and bottom of the climb from some other source known to be more accurate -- then use that to verify your current data set.

problems with using a cyclometer

A cyclometer is a speedometer / odometer designed a bicycle -- and some cyclometers also have an altitude sensor.

This can be the best method -- if used correctly -- because it's closest to the real road and the real riding of it.

Problems to watch out for:

  • odometer / distance funtion not calibrated accurately for the current tires on the bike. This is completely fundamental for measuring steepness -- but you have to find a carefully measured kilometer or mile section of road and actually verify that the cyclometer is measuring accurately.

  • reliability or relative accuracy of the barometer / altimeter sensor for most cyclometers is not well known. I would not guess that accuracy of the altimeter is a high priority for manufacturers of most cyclometers.

  • if a weather front or storm is approaching, a barometric altimeter might not be accurate for a long climbing section. (but I doubt this is usually a major problem for measuring steepness of each specific shorter section of a climb). It's a good idea to have a measurement of the difference in altitude between the top and bottom of the climb from some other source known to be more accurate -- then use that to verify your current data set.

  • modern digital cyclometers use various numerical "algorithms" for averaging and smoothing and "cleaning" the raw data from their analog sensors. On a steep road with variations, these algorithms could be a problem, because (a) the manufacturer has likely "tuned" the algorithms for flatter terrain and higher speeds, and because (b) the algorithm for "smoothing" distance and the algorithm for "smoothing" altitude measurements might interact in funny ways with steepness calculations. Then since the algorithms are usually not well documented, you don't know how to compensate for them.

  • some cyclometers might be deliberately designed to distort altitude or steepness numbers displayed -- in ways which the manufacturer hopes will please the customer. Many customers are pleased to believe that the steepness of a hill is greater than it really is -- because it makes them feel their achievement in climbing it was greater. Very few bicyclists complain to the manufacturer when the instrument tells them their legs are stronger than they had expected.

  • of course if you weave from side to side while you ride up the hill, the distance recorded by cyclometer will be longer, and then the steepness you calculate afterward will be lower. Which is more accurate for your actual riding, but less accurate for measuring the steepness of the overall road.

problems with steepness measurements from websites

The big advantage of a steepness profile from most websites is that you didn't have to do the work yourself.

The problems with most of them are that:

  • You have no clue which measurement approach was used -- topo software? Google Earth? ordinary GPS? GPS with a barometric altimeter? cyclometer?

  • You have no clue what measurement procedures were used -- how was the instrument calibrated? what's the characterization of their inaccuracies? were the measurements repeated a second time?

  • Most bicyclists who put data up on a website have no awareness about how measuring steepness can get so tricky.

  • Some websites seem to be mostly aggregators  (e.g. ClimbByBike) of data from other sources and from lots of different riders, so the accuracy of the steepness data on that sort of website could vary widely -- because carefully verifying the accuracy of each steepness profile submitted is hard work -- and can offend people who try to submit data and report about interesting climbs.

  • Most websites don't have a sound process for other riders to criticize and help improve the elevation profiles and steepness data given.

Suggestion for testing the accuracy of a climb on a website:

Find a discussion forum with local riders who are interested in that hill, try to post to the forum a link to the steepness profile graph data you found on the web, and ask for comments.

Lots of local riders may not be able to do a good job of creating a steepness profile for a climb. But if they're done the climb a few times they can look at somebody else's steepness profile and say how it's wrong.

signs alongside the road

Sometimes there will be a sign along the road giving a numerical percentage for steepness, like 10% or 14%, or once I saw a sign saying "25% grade".

To me it seems clear that road authorities have reasons to post exaggerated steepness numbers, and little incentive to post more accurate (lower) steepness grade numbers.

My experience with such signs in several USA states and European countries is that: (a) after riding or driving the hill, usually I am convinced that the percentage grade number on the sign was greated than the steepness of any significant section of the hill; (b) sometimes I'm not convinced that the number on the sign is somewhat exaggerated, but I sorta doubt that it's correct; or (c) in one case I could name (maybe two?) I have ridden the hill multiple times and calculated steepness with softward, and become convinced that the number on the sign is correct.

Just because you survived climbing up a hill that had a sign that said "14% grade" does not mean that you can handle a hill which really is sustained at 14% steepness.

the real test

Get out and ride up that hill.

Then ride it again.

Sharon + Ken in southern Germany


Overall we had a fun time riding in a wide variety of places. Made us think we'd like to come back to southern Germany again to try some more riding. As usual, we used a rental car to get us and our bike to the different areas we rode in. Reports below for:

see also:

Resources we used:

  • box of 30 maps 1:75000 of different areas: Die schönsten Radregionen. Kartenset mit 30 Einzelkarten (BVA Bielefelder 2006)

  • guidebook: Die 100 schönsten Radtouren in Deutschland (BVA Bielefelder 2007),

  • various ADFC bike route maps (around 1:75000) for local areas

  • detailed city-and-surrounding region road+street maps (esp for Nürnberg+Erlangen, Munich, Frankfurt)

  • book of Michelin 1:300000 maps for Germany + surrounding countries.

We didn't follow any of the routes in the guidebook or maps exactly, but they were very good sources of ideas for our style of riding.

Here's some rough notes as best I can remember from various rides:

Munich city from east

We rode in from Feldkirchen during morning rush hour. Then rode on streets + sometimes sidewalks around obvious tourist sites in the central city (many weren't open yet in the morning). Then north thru park and along the river (some bike paths were packed dirt), east back to Feldkirchen. We liked the city a lot, the riding worked well for us [ photos | very rough map ]. Amazing number of people using bicycles for transportation. 

around the Forggensee

Great ride (but short) around a pretty lake: views of water and mountains, views of animals, variety of terrain (some gentle, some hills, some flat, short section on dirt) [ photos | rough map ]. Got the idea from 100 schönsten Radtouren book. We rode it in the clockwise direction, which worked fine (perhaps the climbs were steeper than if had taken counter-clockwise, but I think the downhills were more fun going clockwise).

Our visit made me think I'd like to try more riding in that southwestern part of Bayern (find more pleasant roads with animals out visible). We also hiked up to the Neuschwanstein castle and higher up to Marienbrücke view -- felt it was worth doing once every 10 or 20 years -- riding around the Forggensee I'd gladly do much more often.

Peterhausen - Dachau loop (NW fr Munich)

Pleasant pretty farming area [ photos | rough map ]. Straightforward navigation except tricky thru the city of Dachau. Mostly moderate, but at least one steep-ish hill. Wished it had more animals out visible from the roads.

around Nürnberg

Great variety, mostly on roads and streets and paths that were good for our kind of riding [ photos | rough map ] - (except never again the cobbles thru the village of Buch). We used ideas from the 100 schönsten Radtouren book and bike maps and a detailed Nürnberg road + street map. We started in Erlangen, rode south into the city of Nürnberg -- walked our bike (often on cobbles) a lot around to visit some different sights of the city. Then south to the Ludwig - Donau - Main canal, east on a packed-dirt path alongside the canal (quiet + pleasant), then north thru Feucht back to the city of Nuernberg. It rained in Feucht, so we decided not to try our original plan of taking another bicycle path along the river west from Nuernberg, and instead more or less retraced our route back to Erlangen. But we ran into some bad problems with our rear tire toward the end, so we arrived back rather late.

Erlangen northwest to ponds + bierkellern

A cousin of mine lives near Erlangen, and she and her family took us out for bicycle ride on a Saturday. First west on pleasant bike paths and streets to Büchenbach, then on some dirt bike paths thru the pleasant woods to some quiet fish ponds. Only problem was that the paths had some short sections not as smooth as the canal paths we had ridden on other days, and the narrow road-riding tires on our tandem got a couple of flat tires -- so we fixed them and continued the pleasant ride. Next on some roads to a bierkeller near Hemhofen -- which turned out to be closed on Saturday (open Sundays), so we ate a picnic lunch on one of the tables, then rode on roads to another bierkeller beside a lake near Dechsendorf, which was open, so we had coffee + cake and sat and talked, and then Sharon and I rode on some mostly pleasant roads back to Erlangen [ photos | rough map ].

Würzburg + southeast

Pleasant visit, first walking around the city and seeing the inside of the famous Residenz, then a short bicycle ride:  up onto a ridge southeast of town to Gartenstadt Keesburg with a view, down thru Gerbrunn to Randersacker, finally a paved path west along the Main river back to the city  [ photos | very rough map ]. We didn't have a detailed road+street map of the city its surrounding area, so navigation for bicycling got tricky, but somehow worked out OK -- we started following the idea from the 100 schönsten Radtouren book, but when it looked like it led into a steep narrow unpaved road, we went a different way that stayed on wide well-paved roads. 

Frankfurt city from the east

We rode into Frankfurt from Offenbach on Sunday. First a pleasant path alongside the Main river, then on the streets on the north side, then return on the bike path on the south side of Main river. Definitely liked it [ photos | very rough map ]. Ken had biked or skated in Frankfurt city twice before, had no hesitation about doing it a third time. City was kinda quiet, might have been interesting to see how it was different on a mid-week day.

Ludwigshafen - Speyer

park Waldsee, ride N to south docks of Ludwigshafen, then S along Rhein river to Speyer, then N back to Waldsee. Idea from 100 schönsten Radtouren (? "Blaue Adria"). Pleasant gentle ride, glad to do it once [ photos | rough map ] - (but it helped confirm that Sharon and I are not excited to try riding day after day on gentle/flat bike paths along a river).

non-bicycling day: climb the Alpspitze

see more current version

One day we decided to try climbing a mountain by a via ferrata route  [ photos ] -- an exciting adventure (for which it was good we'd had some previous European mountain climbing experience). A via ferrata is a climbing route on rock with steel cables on some of the very sections, which we could attach ourselves to for protection against taking a long fall -- we each had a special via ferrata kit to use to do the attaching. English and French speakers usually use the Italian word via ferrata for this, but the normal German word is klettersteig). We got the idea for trying this from a guidebook (in English translation) for klettersteig routes in southern Germany and nearby Austria (of course there's also at least one German-language website for this area).

The Alpspitze mountain peak (elevation 2628 meters) is next to Garmisch-Partenkirchen, a ski resort area south of Munich. It's on the same ridge as the Zugspitze, the highest mountain in Germany. We went up on the Alpspitzbahn lift (and also took it back down again at the end of the day). It was a long climb for us (not practiced for that sort of thing) -- around 600 vertical meters or 2000 vertical feet of climbing. I though it was interesting climbing -- Sharon and I used to do technical rock climbing -- I'd say mostly "class 3" by American standards if use the metal cables and rungs, but perhaps some sections harder. Much of it is on a big face -- I felt like I wouldn't have known where to climb on it if the route weren't there.

Sharon and I felt the hardest part was after reaching the summit -- descending the east ridge on steep scree (loose rock). Then we left the hiking trail and found the old Nordwandsteig route (another via ferrata) and descended that (which worked fine -- sorta interesting, but nowhere good as the newer via ferrata route we had earlier climbed up) to go back to the top of the Alpspitzbahn lift.

see also:  map: overview for trip | photos for overall trip

Tony + Ken in northeast Italy


It was great to do a range of different mountain road rides, and for variety a couple of days of climbing (a kind that was new for both of us). Some of the rides and climbs still had snow visible in June this year, and that made them look more pretty than I remember last time in September. (I think there was a lot of snow this year in April, so in other years there might not be as much showing in late June).

see also:

Overall I thought the three days of riding in Italy were spectacular interesting loops I'd do again, and I was glad to succeed on the challenge of the Grossglockner highway (which I couldn't assess for views because it was a misty rainy day). I also liked the two climbing days in between, and I'll be glad to try some more "via ferrata" climbs in the Dolomites sometime.

Stilfserjoch / Passo dello Stelvio + loop over Umbrail

see photos | see route on map

Tony saw we had good weather to go for Stelvio the day after we got off the airplane. I had been thinking it would be better to get acclimatized to altitude (since it's the second highest paved-road pass in Europe) -- but I decided to go along with his enthusiasm, so we got up early and drove a couple of hours to Prad am Stilfserjoch at the bottom of the climb up the east side of the pass.

"Stilfserjoch" or "Stelvio"? My understanding is that the east side of the pass is in an official German-speaking area of Italy called Südtirol (or "Alto Adige") - (and lots of the riders who climb it are visitors from German-speaking countries) - while the west side is in Lombardia which is officially Italian-speaking. So this pass properly has two different names in two different languages. My preference is to respect this complication by calling it "Stilfserjoch" when climbing it from the German-speaking east side ("joch" is a German word for mountain pass) and "Passo dello Stelvio" when climbing it from the Italian-speaking west side.

We just parked in the main parking garage in the village and rode a short ways to the main road and just started climbing. Tony soon left me behind, since my bike has a third small chainring and his doesn't, so there was a limit to how slow he could go even if he wanted to try to ride together with me -- anyway he was just faster. It was so great to be there on a sunny day with snow visible on all the surrounding mountain peaks -- so I stopped lots of times to take lots of photos.

Last time a couple of years ago I climbed and descended the east side on a cloudy day with much less snow on the peaks mostly in the mist, and I was unimpressed. But this time I thought it was the most spectacular paved-road mountain pass I've climbed on my bike.

So for me it was really about the wonderful mountain scenery. Main impressions of the riding itself:

  • The climbing just goes on and on

  • Lots and lots of motorcycles (more than cars, more than bicycles)

  • I was careful to go slow + easy

  • I actually passed two other bicyclists (unusual occurrence for me on road-bike climbs)

  • I felt strong coming to the top.

Tony and I had a snack at one of the restaurants at the pass, then rode down the west side to below the middle set of switchbacks (which are just above the five tunnels), then turned around and climbed up thru the switchbacks.

The scenery on the west side was not as spectacular as the east, but plenty good enough. I felt the riding itself had more interesting variety on the west side, and I was glad the descent was mostly less steep, because for me that tends to be more fun because I don't have to do so much braking.

But we did not climb all the way up to Stelvio pass again. Instead we turned north to go over Passo Umbrail into Switzerland, and descended that a long, long way (partly on packed dirt) often steep to Santa Maria in Val Mustair, then southeast down moderately in Val Mustair, into Italy at some point and east to Glurns (or "Glorenza"),  and south on a gentle-to-moderate road back to our starting point in Prad am Stilfserjoch.

I didn't find the north side of Passo Umbrail as spectacular as either side of Stelvio / Stilfserjoch, but much quieter, and then lower in the Mustair valley was pretty farmland, so I was glad for the variety, and I would ride this loop again the same way.

Sella Ronda - longer variation

see photos | see route on map

We rode on the Sella mountain group (Italian "Gruppo di Sella") in the counter-clockwise direction, but substituted Passo Falzarego and Passo Valparola for Passo Campolungo.

Sharon and I had ridden the normal shorter loop of Pordio + Campolungo + Gardena + Sella on our tandem three years ago [ photos ], and then I rode a longer loop [ photos ], but this day was the first time I tried the climb up the southwest side of Passo Falzarego.

I loved the descent east from Passo Pordoi to Arabba, and we stopped for a snack. Descending further east from Arabba was pleasant, then the climb up the southwest side of Falzarego was pleasant and mostly not real steep, but to me didn't feel as spectacular as some other climbs in the Dolomites (though the top part was pretty good). Tony as usual climbed faster, and we had a snack together at the pass. A little higher to get over Valparola. I found the descent of the north side of Valparola a little steeper than is fun for me, but Tony said he liked it -- then I had more fun on the lower section. Gentle-moderate up to Corvara and a stop for ice cream.

Climb up the east side of the Passo Gardena (German "Grödnerjoch") was spectacular as always. Descent to southwest felt OK to me, but not great. Climb up to Passo Sella felt hard (after climbing both sides of Stelvio the day before, and three other passes already today) -- and then it started raining just after we reached the pass. I didn't enjoy the descent much (steep and wet road), but Tony said he liked it.

Next day we were driving over Sella and Gardena in the rental car to reach a climb, so then we took photos in more favorable light.

We rode the loop in the counter-clockwise direction (seemed great to me), but I noticed that lots of other riders were riding the normal shorter loop in the clockwise direction. What I like about counter-clockwise: Pordoi east descent, Campolungo north descent, Gardena east climb is spectacular and not real steep, avoids the longest steepest climb (up south side of Sella, hard for Sharon + me on our tandem) -- but I'm not so much a fan of descending south side from Sella. What I like about clockwise direction: descent east from Val Gardena, descent north from Sella -- and I haven't had a chance to try the descents of south side of Campolungo and west side of Pordoi.

Next day we climbed the highest peak in the Sella group.

around Monte Cristallo near Cortina d'Ampezzo

see photos | see route on map

After a long day of climbing (coming after two days of riding), Tony and I were ready for a shorter easier day of riding. So we tried riding around Monte Cristallo, a big peak northeast from Cortina d'Ampezzo.

So if riding around the Gruppo di Sella is called the "Sella Ronda", seems to me this loop should be called the "Cristallo Ronda" (but I've never heard that name).

We started in Pocol, above the west side of the famous ski resort of Cortina d'Ampezzo. First descended to Cortina (good views, good riding). Rode around town a little, then headed north, soon found a nice bike shop "Cicli Cortina". Continued riding north on the SS51 toward Toblach / Dobbiaco, mostly moderate and pleasant, less vehicle traffic than I expected, occasionally some nice views of peaks. Met a couple of Czech riders, down a little to the Schluderbach / Carbonin junction . . .

Then a side trip to the Dürrensee / Lago di Llandro. I somehow remembered (from a day of cross-country skiing) there was a view east from there to the Tre Cimes peaks, but we didn't find it from any paved road. What Tony did find by walking a few steps east from the road was a great view of Monte Cristallo looking south across the Dürrensee lake.

Next we climbed up to the lake Misurina -- some steep sections. Along the way I saw the turn for the steep climb up to Rifugio Auronzo, but we weren't up for it on that day. Misurina had nice "lake with mountain" views, and we refilled water battles at a trough on its north side.

Some ups and downs to Passo Tre Croci, then a long long descent west to Cortina, with some nice views. I liked it overall, some parts of the descent were steeper than was fun for me -- but Tony said he liked the descent a lot. A gelato stop on the main street of Cortina, then the climb back up to Pocol seemed not too steep and not too long.

Overall I would gladly ride that loop again, and in the same direction -- because I felt the SS51 road around the west side of Cristallo was not curvy enough to be an interesting descent, but it would be a doable climb for Sharon and me on our tandem, then we might if necessary walk up some of the steep sections around the east side of Cristallo.

Grossglockner highway north side (in Austria)

see where on map

On our last day I ride up the north side of the Grossglockner highway and finished on a higher spur road on the Edelweißspitze (2571m). The Grossglockner peak is the highest summit in Austria, and the highway goes over passes on its east side. The highway is a toll road, but they didn't charge me to ride my bicycle on it.

I had seen a steepness profile on the web which showed the north side as having a long section (about 380 vertical meters) around 11% steepness grade, including with one section steeper. That was steeper than any long climb I'd done before, so I doubted I could do it. But my long-time ski partner Gi said he knew of some non-super-athletes who had done it. So with that encouragement I drove very early in the morning from Salzburg down to the village of Bärenwirt (a bit south above Fusch), and started climbing around 6:30.

Some pleasant views across the valley, but unfortunately the peaks were mostly in the mist. I was glad to find out that the climbing was well within my capability. I made about three pauses for snack +  drink during the cliimb. I encountered two other bicyclists who were coming down already. A short ways before reaching the FuscherTörl (a pass on the north side at 2428m), I turned sharp left onto a spur road, and climbed that (with some cobbles) to a parking area at the top of the Edelweißspitze (2571m) -- but I couldn't see any peaks thru the mists.

On my descent I encountered another rider climbing on a arm-powered recumbent tricycle (I believe he climbed up the entire north side highway). Next I visited the FuscherTörl, turned around and started back down. Soon it started raining. Even with wind jacket and wind pants, I was shivering -- and the descent felt too steep to be much fun for me. Lower down I saw lots of riders on mountain bikes climbing up -- lots more mountain bikes than road bikes. (Perhaps the point is that the lower gears make the steep climbing easier, which the knobby tires help slow the descent.)

Next time I think I would wait for a sunny day when I could see the views, and earlier in the season with more snow, then ride higher to see more around on the southeast side -- or try climbing from the south side in early season before most of the sun-exposed snow melts away.

non-riding day: "via ferrata" climb on Piz Boe

see photos | see where on map : see more current version

After two long hilly days of riding, we decided to try something different. The Dolomites are famous for a different kind of steep rock climbing -- where the are steel cables and bars ("ferrata" = iron) attached to the rock to protect against taking long falls (but not necessarily protect against a shorter fall). Also sometimes help get up thru difficult sections of climbing. We'd never done it before. We had a via ferrata "kit" of special protection equipment and helmet and shoes good for climbing on rock for each of us, and a guidebook and some hiking maps.

We started in Corvara, took the Boe and Vallon lifts as high as they went -- into the heart of the Sella mountain group -- which we had just ridden around on our bikes the day before. We walked following the 646B trail, which later got indistinct and then snow-covered, so we walked up the snow to the left of the waterfall at the head of the valley or ampitheater, at its west end. We put on our via ferrata kits and helmets, clipped into the cable, made rock climbing moves. When our protective attachment reached the next anchor, we clipped into the next higher section of steel cable. But then we reached a section where the cable was buried under the snow, so we unclipped and found a way to stay on easier rock to get to the next section where the cable was on rock, and did some more climbing with protection.

Then a footbridge below a waterfall (got a little wet), then on the other side a steep face -- the most difficult section. Careful moves with a long reach required to clip the next section of cable. Then some steep rock-climbing moves, then less steep and I find myself at the top standing on flat ground in the sunshine. Soon Tony was up to join me.

What next? We decided with all the snow that we did not want to try the shortcut of descending the "Lichtenfelser Steig" (which the guidebook says is path 672 roughly in the northeast direction), because we thought it might have steep sections but with the protective cable buried under the snow. Instead we kept walking up toward the Piz Boe, at 3152 meters the highest summit in the Sella mountain group.

Yesterday we had ridden our bicycles around the Sella group, today we were climbing high inside it.

Navigation was a bit tricky at first because the hiking trail markings were buried under snow, and we could not yet tell which summit was the highest. After we recognized that we had guessed the wrong peak, we corrected course and walked up on alternating sections of snow and rock, until we finally found a marker for path 672 going in the southwest direction. Soon we were up on an interesting rock ridge, then the trail markers down to a snow col at the head of the Val di Mesdi (its south end) -- a long valley that goes down to near Colfosco (just west up the road from Corvara) -- I had looked up the Val di Mesdi while climbing to Passo Gardena yesterday.

Then up on the rock ridge and (roughly) follow along this to the summit -- with a big solar panel -- and hut serving food + drink, called "Capanna Piz Fassa". We each had two strudels and ice tea, and asked the hut warden if path 638 was OK to go back to the Vallon lift, and based on his response decided to take it. Some parts of path 638 had very deep mushy snow which slowed us down (and got our socks soaked). Later a steep southeast-facing couloir filled with snow. Fortunately the snow was still reasonably soft, so we made it down OK facing out down the hill (or sideways) -- but it was strenuous for our leg muscles and joints to have to go down it carefully. 

After that the path was mostly straightforward, but then a thunderstorm hit, so we walked faster (and ran) and then hail started. Tony still had his helmet on from the via ferrata, but I didn't. The hailstones seemed kinda big, so I put both hands on the top of my skull to protect my brain. Then the storm stopped and we arrived at the top of the Vallon lift, and the lift attendant started the chairlift and we rode down on it, then down the Boe ski lift -- and after a long ride we were back in Corvara to finish a fun adventure.

non-riding day: "via ferrata" climb of Cirspitze V

see photos | see where on map : see more current version

For our second "via ferrata" climb, we chose something with shorter access (since Tony's legs were still sore from the long walk down from the top of the Piz Boe). Overlooking Passo Gardena (or "Grödnerjoch") - (where we had ridden over on our bikes three days earlier) - is a row of peaks, and the fifth one is called the Cirspitze V (older spelling "Tschierspitze"; Italian "Piz da Cir V"). Nice sunny day, pleasant hike north from the pass, then some rock scrambing already before reaching the "iron" section. One ladder at the start and the rest of the climbing was on rock, with the steel cable for protection. A couple of times I grabbed the cable to help with a harder move, but the rest of it I enjoyed using only the natural rock features -- interesting climbing moves -- exciting in places, even though I knew I wouldn't take a long fall.

A couple of times beginner climbers far above me dislodged pebbles that fell very close to me, so I was very glad I was wearing my helmet.

The summit was pretty small, so after declaring victory at the top we find another nice spot to its north for a little picnic. Spectacular setting, enough different from bicycling.

Upper part of the descent gully had a cable for protection -- which both of us also grabbed and hung out against to help soften our down-moves. Then a long downward section with some loose rocks with no cable -- so we used hiking poles to help soften the impact on our ankles and knees. Then to a rough trail, then to a packed dirt road and back down moderately on it to our car parked along the main Sella Ronda road.

see also:  map: overview for trip | photos for trip

estimates for mountain rides in northeast Italy


These are some rough estimates of vertical climbing and horizontal distance for some rides I've considered trying.

what's here:

to change climbing in vertical meters (m) to vertical feet, multiply by 3.3 (or 3.28).

to change distance in kilometers (km) to miles, multiply by 0.6 (or 0.621).

rides near Canazei

around Sella

loop clockwise around the Sella group ("Sella Ronda"): climbing over P Sella, P Gardena, P Campolongo, P Pordoi = +1500m / 53km [ = +5000ft / 33 miles]

longer: replacing P Campolongo with P Valparola + Cernadoi adds +750m for a total of +2250m.

climbing up from Canazei (and finishing there) adds +450m for a total of +1950m / 63km.

longer: starting in Canazei, then replacing P Pordoi with P Fedaia adds something like +600m / 30km for a total of +2550m / 93km.

longer: starting in Canazei, then replacing P Campolongo + P Pordoi with P Valparola + Cernadio + Caprile + P Fedaia adds +1050m for a total of +3000m.

much longer loop: starting in Canazei, then replacing P Campolongo + P Pordoi with P Valparola + Pocol + P Giau + Caprile + P Fedaia has a total of +3700m / 115km [= +12000ft / 71 miles]  (adding a side trip from Pocol to Cortina brings the total to +4000m / 131km).

around Marmolada (3343m)

loop counter-clockwise: Canazei - Moena (1194m) - P San Pellegrino (1918m) - Cencenighe (~775m) - Alleghe - Caprile - P Fedaia (2057+m) - Canazei = +2050m / 96km (? or is it +2250m / 93km ?) [ = +7000ft / 59 miles]

replace Alleghe with around Monte Civetta (3220m): Cencenighe - Agordo (611m) - P Duran (1601m) - Villa (935m) - Forcella Staulanza (1773m) - Caprile:  adds about +1625m / 49km for a total of +3675m / 145km

Fedaia - Giau - Falzarego - Pordoi loop

Canazei (1465m) - P Fedaia (2057+m) - Caprile (1014m) - P Giau (2233m) - Pocol (1453m) - P Falzarego (2105m) - Cernadoi - Arabba (1602m)- P Pordoi (2239m) - Canazei = +3300m

around Sasso Lungo / Langkofel (3179m) + Seiser Alm

loop counter-clockwise: Vigo di Fassa (~1250m) - Canazei - P Sella (2244m) - Santa Cristina (~1500m) - Monte Pena (1635m) - (dirt) - Saltern / Saltaria (~1695m) - Seiser Alm - Compatsch / Compaccio (1844m) - Seis / Siusi (998m) - Tires - P Nigra (1688m) - P Costalunga / Karerpass (1745m) - Vigo di Fassa = roughly +2900m / 112km (making some allowance for dirt)

the dirt could be avoided by going from Santa Cristina to Kastelruth / Castelrotto by way of St Ulrich / Ortisei and Panider Sattel / Passo di Pine -- see below: "loop without Seiser Alm + without dirt".

loop without Seiser Alm + without dirt:
Vigo di Fassa (~1250m) > Canazei > P Sella (2244m) > St Ulrich / Ortisei (~1200m) > Panider Sattel / Passo di Pinei (1437m) > Kastelruth / Castelrotto (1060m) > Seis / Siusi (998m) > Tires > P Nigra (1688m) > P Costalunga / Karerpass (1745m) > Vigo di Fassa = +2600m / 100km [ = +8500ft / 62 miles ]

add down+back to the Eisack / Isarco river valley adds roughly +550m of climbing.

up-and-back to Seiser Alm from west:  Kastelruth / Castelrotto (1060m) to Compatsch / Compaccio (1844m) and Saltern / Saltaria (~1695m) and back down is around +1000m / 34km

one-way Kastelruth / Castelrotto across P Nigra + P Costalunga to Vigo di Fassa = roughly +1350m / 47km

Vigo di Fassa to Canazei add +200 / 11km

Kastelruth / Castelrotto to Canazei total +1550m / 58km

one-way Bozen / Bolzano (266m) across P Nigra + P Costalunga to Vigo di Fassa = roughly +1750m / ??km

Bozen to Canazei total roughly +1950m

around Cima Bocche

loop counter-clockwise:  Moena (SW from Canazei) (1184m) > Predazzo (1018m) > Passo di Rolle (1970m) > low point (?1600m) > Passo di Valles (2033m) > Falcade (1297m) > Passo di San Pellegrino (1918m) > Moena = roughly +2000m / 72km

rides near Corvara + Colfusco

the rides around Sella under "rides near Canazei" pass thru Corvara.

these rides under "rides near Canazei" go one pass-crossing away from Corvara:

  • Fedaia - Giau - Falzarego - Pordoi loop

  • around Sasso Lungo / Langkofel (3179m) + Seiser Alm

rides near Cortina d'Ampezzo

around Cristallo (3221m)

loop clockwise:  Cortina (~1150m) > west pass (1530m) > Schluderbach / Carbonin junction (1432m) > Col St Angelo (1756m) > Misurina > P Tre Croci (1805m) > Cortina = roughly +1000m / 40km [ = +3300ft / 25 miles ]

side trip to Rifugio Auronzo (2320m) and Tre Cime di Lavaredo adds a very steep +600m

around the Tre Cimes / Drei Zinnen

loop counter-clockwise: Schluderbach / Carbonin (1432m) > Col St Angelo (1756m) > Misurina > Auronzo (~770m) > Passo del Zovo (1476m) > Dosatedo (~1200m) > Kreuzbergpass / Passo di Monto Croce (1636m)  > Sexten > Innichen / San Candido (1174m) > Toblach / Dobbiaco (1241m) > Schluderbach (1432m) = roughly +1900m / 93km [ = +6200ft / 58 miles ]

adding the climb in Pragsertal / Val di Braies over Plaetzwiesen / Prato Piazza (1993m)  to Schluderbach / Carbonin adds about +700m / 20km, for a total around +2600 / 113km.

starting from Cortina and going around Cristallo adds about +750m

see also analysis under Auronzo

around Averau (2648m) + Nuvolau (2575m)

loop counter-clockwise:  Pocol (1453m) > P Giau (2233m) > point 1300m > Cernadoi > P Falzarego (2105m) > Pocol = +1600m

starting from Cortina adds +300m

around Antelao (3263m) + Sorapis (3205m) + Marmarole

loop counter-clockwise: Cortina (~1150m) > Pieve di Cadore (~700m) > Auronzo (862m) > P Tre Croci (1805m) > Cortina = +1200m / 93km [ = +4000ft / 58 miles]

add side trip / loop counter-clockwise to Santo Stefano di Cadore (908m) + Comelico + Passo del Zovo (1476m) adds +700m / 30km for a total of +1900m / 123km.

could also add around Cristallo.

around Monte Pelmo (3168m)

loop clockwise: Cortina (~1200m) > low point on SS51 (775m) > Passo Cibiana (1530m) > Forno di Zoldo (810m) > Villa (935m) > Forcella Staulanza (1773m) > Selva di Cadore (1335m) > Passo Giau (2233m) > Pocol > Cortina = roughly +2700m / 97km 

or to do the "whole" Giau west-side climb, descend from Selva di Cadore to Caprile (1014m) [adds at least +350m climbing] - (? is it possible to ride from Selva to Caprile over Colle San Lucia? -- adds more climbing) 

see also

rides near Auronzo + Santo Stefano

around the Tre Cimes di Lavaredo / drei Zinnen

loop counter-clockwise: Toblach / Dobbiaco (~1200m) > Col St Angelo (1756m) > Auronzo di Cadore (~850m) > #532 north over Passo del Zovo (1476m) > Comelico (~1180m) > Kreuzbergpass / Passo di M Croce di Comelico (1636m) > Innichen / San Candido (1174m) > Toblach / Dobbiaco = +1800m / 80km (or is it 93km -- see above under Cortina) [ = +6000ft / 50 miles]

side trip to Rifugio Auronzo (2320m) by Tre Cime di Lavaredo adds a steep +600m.

replace Col St Angelo with Cristallo Ronda + Cortina loop counter-clockwise adds +500m / 33km for a total of +2300 / 113km.

start from Cortina is same as "replace Col St Angelo" [ = +7500ft / 70 miles]

see also analysis under Cortina

Auronzo loop east to Comeglians (in Fruili)

loop clockwise: Auronzo (862m) > #532 north over Passo del Zovo (1476m) > Comelico > SS52 southeast to Santo Stephano (908m) > #355 east thru Sappada > over Cima Sappada (1286m) > Comeglians (553m) > #465 west to Forcella Lavarder (1542m) > #619 west over Sella di Razzo (1760m) to around Vigo + Lozzo (~700m) > north to Auronzo = +2500m-2700m / 114km [ = +8500ft / 71 miles]

add Comeglians - Ravascletto - Sella Valcalda (959m) side loop adds +400m

add loop Comeglians (553m) > (very steep) Monte Zoncolan (1740m) > Sutrio > Paluzza (~560m) > Sella Valcalda (959m) > Comeglians adds +1600m.

add Comeglians - Tolmezzo loop (clockwise): Comeglians (553m) > Sella Valcalda (959m) > Paluzza - south thru Avosacco > Terzio > Tolmezzo (323m) > Moia > Comeglians adds +500m / 54km.

see also

rides near Toblach / Dobbiaco

around the drei Zinnen / Tre Cimes di Lavaredo

see under Cortina or Auronzo

Austria mountain loop (around the Deferregeralpen -- Weissespitze 2963m)

loop clockwise: Toblach > Anthölz valley > Stallersattel > Lienz

Toblach / Dobbiaco (1241m) > SS49 west down (1015m) > Anthölz / Antarselva > Stallersattel / Passo Stalle (2052m) > east into Austria > Huben > rt 108 southeast to Lienz city (673m) > Sillian > west into Italy > Innichen / San Candido > Toblach (1241m) = roughly +1800m / 146km.

around the Lienzer Dolomiten (all in Austria)

loop counter-clockwise: Sillian (1102m) > SE on rt 111 to Kartitscher Sattel (1529m) > Obertilliach > E on rt 111 to Koetschach-Mauthen (708m) > climb N on rt 110 over Gailburg Sattel (982m) > Oberdrauburg (621m) > W to Lienz city (673m) > Sillian (1102m) = roughly +1200m / 121km.

rides near Stelvio + Bormio

Passo dello Stelvio / Stilfserjoch (2758m)

West side from Bormio = +1550m / ?? km.

East side of Stilfserjoch / Stelvio:

Franzenshöhe (2189m) up to Stilfserjoch (to do the final 20 or 22 switchbacks above treeline with views of snow-covered peaks = about +600m [ = +2000ft]

Ristorante Rocca (1861m) up to Stilfserjoch = about +900m

Trafoi (1547m) up to Stilfserjoch (to climb 46 out of the 48 switchbacks) = about +1200m

Prad am Stilfserjoch / Prato allo Stelvio (913m) up to Stilfserjoch = about +1800m [ = +5900ft]

start Bormio, climb up W side of Stelvio pass, over to E side to Franzenshöhe and back = +2150m

Bormio  up W side of Stelvio pass, over to E side to Trafoi and back = +2750m

loop clockwise with Val Mustair:

Prad am Stilfserjoch / Prato allo Stelvio (913m) > Trafoi (1547m) > Stilfserjoch / Stelvio (2758m) - Passo Umbrail (2498m) > descend (including 3km of dirt-gravel) to Santa Maria im Mustair (1375m) > gentle in beautiful Mustair valley > Glurns (Glorenza) (908m) > Prad am Stilfserjoch = +1900m / 62km [ +6200ft / 38.5 miles]

side trip to Pass dal Fuorn / Ofenpass (2149m) adds +800m for a total of +2700m

side trip to Pass dal Fuorn to Zernez CH (1472m) and back adds +1500m for a total of +3400m

loop from Bormio to Passo Umbrail + Val Mustair + Stelvio:

Bormio (1217m) > climb to Passo Umbrail (2498m) > descend (including 3km of dirt-gravel) to Santa Maria im Mustair (1375m) > gentle in beautiful Mustair valley > Glurns (Glorenza) (908m) > Prad a. Stilfserjoch / Prato allo Stelvio (913m) > Trafoi (1547m) > Stilfserjoch / Stelvio (2758m) > descend to Bormio = +3500m [ = +11500ft]

Passo Gavia (2621m)

Gavia -- Passo Gavia (2621m) from North side from Bormio (1217m) = +1400m [ = +4600ft]

South side is steeper, and has narrow sections and an (avoidable) dark tunnel.


around Monte Civetta (3220m)

loop counter-clockwise: Agordo (611m) - P Duran (1601m) - Villa (935m) - Forcella Staulanza (1773m) - Caprile - Alleghe - Cencenighe -  Agordo = +1625m / 73km [ = +5300ft / 45 miles]  

more . . .

see also


concept words: roberts Europe European visit trip vacation holiday American visitor report reports

bicycling: bicycle bicycling bike bikes bicycles bicyclist cycle cyclist cycling touring riding rider riders

routes: route routes ride rides tour tours map maps

places: place river valley state country region regions area areas city town village

fahrrad rad radfahren radtour touren routen velo tour tours route routes velotour cyclisme

bici bicicletta percorso itinerario visita giro turistico

Italia Italie Deutschland Allemagne Alemania Germania hautes cols alta alto höhe