Winter route design: ice, shade and the roads that thaw first

Winter riding is not mainly a clothing problem. Clothing is solved; the kit works, and most riders sort it out in one season. What stays unsolved is the route, and the route is what actually decides whether a December ride is pleasant, dull or a slow slide into a hedge.

The useful thing about winter road hazards is that they are highly predictable in space. Ice does not appear at random. It forms where the road loses heat fastest and gains it slowest, and both of those are functions of terrain you can read off a map before you leave. This post is about how to do that reading.

Aspect is the biggest single lever

At 50° north in late December the sun climbs to around 16° above the horizon at noon, and it is barely above the horizon for most of the short day. At that angle, the amount of direct radiation a road surface receives depends enormously on which way the ground faces. Topographic solar radiation models — the ones used in GIS to compute incoming radiation across terrain — show slope and aspect dominating the result at high latitudes and low sun angles.

The practical version: a road cut into a north-facing slope may get no direct sun at all between November and February. Whatever froze overnight is still there at midday. The road on the opposite side of the same valley, facing south, has been in weak sun since mid-morning and is dry.

You can see this on any map that shades relief, and you can see it more precisely on any planner that shows you which way a road runs relative to the contours. It is the first thing to check on a winter route, and it applies most sharply to descents, where a rider is fast, cold and has the least ability to correct.

Frost hollows are real and they are local

On a clear, calm night the ground radiates heat to a cold sky, the air immediately above it cools, becomes denser, and flows downhill. It collects in valley bottoms, hollows and behind embankments. Road meteorologists have been measuring this for decades with instrumented vehicles, and the results are consistently striking: road surface temperature can vary by several degrees Celsius between points only a few kilometres apart on the same route, on the same night.

Two conditions govern how strong the effect is. It needs clear skies, so the ground can radiate freely. And it needs light winds, because wind mixes the air layer and flattens the differences out — measurements taken on clear but windy nights show much smaller variation along the same routes. That gives you a usable forecast rule.

The dangerous morning is the clear, still one after a cold night, not the windy one. A blustery night at the same air temperature is far less likely to have produced patchy ice in the dips, because nothing had the chance to settle and pool.

Shade: more complicated than it looks

The received wisdom is that shaded roads are icy roads. That is broadly right during the day and slightly wrong at night, and the distinction is worth having.

Overnight, a road screened by hedges and trees has less open sky to radiate to. Studies measuring the effect of screening on road surface temperature find that sheltered sections cool more slowly than open ones for exactly that reason, so the open field-edge road can actually be the first to reach freezing. During the day the relationship reverses: the screened section gets no sun, so once ice has formed it persists for hours after the exposed road has cleared.

The route-planning consequence is that shade is not a hazard in itself — it is a persistence multiplier. If nothing froze, shade costs you nothing. If something froze, shade is why it will still be there at two in the afternoon. That is the case for riding the woodland lanes later in a cold spell and the open roads earlier.

Gritted roads and the honest trade-off

Highway authorities have a statutory duty to ensure, so far as is reasonably practicable, that safe passage along a highway is not endangered by snow or ice — and they discharge it by treating a defined priority network rather than every road. Main roads, bus routes and approaches to hospitals get salt. The pleasant lane over the hill does not.

This creates a direct conflict with the usual advice, and with the thinking in seven rules for building a good route, which pushes you towards quiet roads with few junctions. In winter, the quiet road is the untreated one. There is no clean answer here, and anyone who offers you one is selling something. What is defensible is making the trade deliberately: on a genuinely icy morning, a treated B-road with more traffic may be the lower-risk option than an untreated lane, and on a mild damp day the reverse is obviously true.

Research relating slipperiness type, maintenance activity and crash risk finds that the treatment does matter — and also that the riskiest conditions are the ones that arrive unexpectedly, such as hoar frost or rain falling onto a frozen surface, rather than the obvious snow everyone prepares for. Most authorities publish their gritting route maps. Looking at yours once, in October, is twenty minutes that pays back all winter.

The hazards, in the order they appear

Daylight is a budget, not a hope

In late December at 52° north you have roughly eight hours between sunrise and sunset, and the first and last of those are dim. Riding in darkness carries a measurably higher crash risk than the same road in daylight, and the estimates from studies that control for the confounds are not small.

Treat daylight as a hard budget. Work out the time from usable light to usable light, subtract a buffer for the punctures that take four times as long in cold hands, and size the ride to what remains. If the answer is 2 hours 40, plan a 2 hour 40 ride rather than a three-hour one you intend to hurry. If you do end up finishing in the dark, riding at night covers the lighting properly.

Shorter loops, more of them

The last piece of winter route design is structural. Replace the one long loop with a cloverleaf: two or three shorter loops that all pass back through home or through the same village. You get the same distance, and at no point are you 45 km from a warm room with numb hands and a route that only goes forwards.

This also solves the abort problem. A big loop commits you at the halfway point; a cloverleaf lets you stop after one leaf without failing at anything. In practice riders finish more winter rides this way, because the decision to continue gets made three times when you are warm rather than once when you are not.

And on the mornings when the roads genuinely are not rideable — freezing rain, sheet ice on the priority network, the kind of conditions where skill does not help — that is what the trainer is for. Moveee Indoor will ride one of your own saved routes in a browser tab, which at least means the session you lose outside is the session you do inside. Between that and the road, most winters have very few days that need to be written off entirely. For the training side of the season, winter training for cyclists and how to dress from 0 to 15 °C cover what to do with the hours once you have found the roads.