I’m looking for input on using Key:ice road for street vehicle traffic. There is not very much described in the wiki on how to map. I tried looking around for a good example in overpass turbo but came up empty handed since most ice roads are mapped for snowmobiles.
The road in question is informal. It is plowed each year and kept open by locals. It is specifically for street traffic to cut off 30 minutes of their drive around the lake. There are obviously no rules as to what vehicles can go on it. Do the current tags on Onigum Cut Across look right? Anything I should add or remove?
I originally had it as an unclassified road because it is normally very wide and smooth. It is not an official road and I was nervous about routers picking it up in fall/spring when the ice is unstable. So I downgraded it to a track.
I suspect highway=track would dissuade car routing profiles from voluntarily taking this shortcut unless forced to, but pedestrian routing profiles would still take it.
Unfortunately, I don’t think any of the mainstream routers understand seasonal=winter. There’s an open feature request about that for Valhalla. GraphHopper has talked about implementing seasonal restrictions, but seemingly not the case you’re describing. To be sure, you could tag something like access=noaccess:conditional=yes @ (Nov 1 - Mar 1), but you’d have to come up with a specific date range during which you’re confident that the road is passable.
Recently I’ve been working on improving documentation (and investigating) winter sport “pistes”, which often reuse roads, parking lots and other facilities. That creates a complementary problem – how to express that the cars may not use the facility in winter.
I ended up tagging this parking lot, which I used as an example in piste:type=ice_skate, with the conditional tag motor_vehicle:conditional=no@Dec-Mar, but I’m far from sure if that’s a viable solution.
I wonder if ice roads should even use the highway=* key at all. In some jurisdictions, ferry routes are officially designated part of the highway network, but we tag them asroute=ferry, ferry=primary not highway=primary + ferry=yes.
Meanwhile with this ice road tagged highway=primary + ice_road=yes any renderer that doesn’t modify their standard highway rendering when ice_road=yes is present (most of them), displays it in a way that looks like it might be a causeway across the ocean:
This is in the Gulf of Anadyr off the Bering Sea near the Arctic circle, so it is typically ice covered 10 months out of the year, but it is still quite a different type of highway than those built on land.
But where do we draw the line? If it’s a road on land, but impassable during wet season, but it’s still an important regional road because there aren’t any better all-season roads, can it get highway=primary or no?
Yeah the line could be a little tricky to find. Distinguishing ice roads built over frozen water from roads built on land would seem fairly straightforward, but what about an ice road built across a wetland? Is that over land or water? Thankfully I think this question doesn’t really need to be answered because ice roads aren’t just normal roads that happen to have a surface of ice. Although commonly built over water, they are built over land as well. Their defining character is that they are re-constructed each winter from ice and snow rather than being permanently constructed from materials like dirt, gravel, asphalt.
A dirt or gravel road that is impassible during the wet season is likely still recognizable as a road during that time. Even if it gets entirely submerged in water, the roadbed still exists beneath. On the other hand, it sounds like ice roads mostly disappear once the snow and ice melts because those are the materials the road bed is constructed from. For ice roads over land, there is no roadbed on the ground below the snow and ice. For ice roads over water there is no solid ground at all. In this way they are like ski and snowmobile trails that exist only in winter.
Of course this video also explains how climate change is threatening the viability of building ice roads. So maybe if we just wait a few more decades we won’t need to think about how to tag them anymore.
In my point of view, if it’s a road (seasonal) why should it not be a normal highway=* based on it’s importance? Together with surface=ice and some kind of opening_hours-description on when it can be used.
Finland has three ice roads (for normal road traffic but with weight restrictions) that are part of the official road network / maintained by a governmental agency. One of them is not mapped in OSM, the other two are:
The tagging seems to be a bit shaky. I think Finnish practice is to not use a highway tag for snowmobile routes (when crossing waters or otherwise, a suitable value doesn’t seem to exist.)
For unofficial roads I’d stop to consider the implications of tags like highway. E.g. Would that that imply some kind of traffic law applies?
Depends on which aspect of traffic law, I suppose. Minnesota (where Onigum Cut Across is located) apparently allows motor vehicle access to iced-over lakes and waterways by default unless otherwise restricted. Staying on the ice road is recommended but not strictly required (correct me if I’m wrong). However, there are weight restrictions and presumably the usual rules about how to operate a motor vehicle.
Yes, and the term winter road is used in standard English as well. The meaning overlaps with ice road. All ice roads are winter roads, but not all winter roads are ice roads.
I agree with this. In winter, these are solid roads and for some even for heavy trucks. OSM is a database where we document these highways with various characteristics. This is to Map Styles and routing software to adapt to this reality.
The link below ia part of some 500km long ice road that I traced a few years ago is in a region with no roads and accessed by ferry. The ice road is maintained in the winter season for snowmobiles and is quite a great moment for villagers and tourists. It moves over land, often wetland, and sea. Believe me, this is a solid road in winter, in particular this year. Yes winter is back !
Tagging ice roads this way is not entirely unreasonable (it’s what is currently done after all), I’m just wondering out loud whether the nature of ice roads is different enough that they could be deserving of their own top level tags. Unlike other road construction materials, the ice these roads are built from melts during the summer. After this happens the road isn’t just unusable, it ceases to exist. When the next winter comes, the ice road is entirely re-constructed from new snow and ice. This is quite a unique yearly lifecycle that most roads do not go through.
Absolutely! Ice roads are serious infrastructure in the north. It’s quite an incredible engineering effort to rebuild them every year once the snow and ice is thick enough.
It sounds similar to back-country (mountain) roads not plowed during winter. Just they get usable when the snow is gone and get unusable with the first snow. Dirt-roads are also frequently “re-paved” by a grader, not always following the original route. I would think putting all those in something new will end up in lot’s of edge cases.
The conventions were established something like 15 years ago and I wasn’t there so I cannot correctly represent everything that was considered. But as I understand it, yeah, ugly on water and downright misleading everywhere else for the rest of the year (i.e. suggesting a path/track exists where there is nothing but forest floor). I believe one of the reasons for not going for highway=snowmobile was that sometimes snowmobile routes overlap existing paths, forest roads etc.
A specialist renderer was created to provide the necessary information to end users (where the routes are, which law applies, who maintains the route, is a permission required).
