Blog

Seasonal factor groups: why they make or break an annualized count

A long, empty two-lane rural road lined with bare trees under a cloudy winter sky.

You pull a 48-hour tube count off a rural collector in October, run the math, and get an AADT. Six months later someone questions the number because the count happened during harvest season on a road that sees a completely different traffic mix in July. This is the problem a seasonal factor group is built to solve, and if you've annualized more than a handful of short counts you've already leaned on one whether you knew the term or not.

What a seasonal factor group actually is

Most state DOTs run a network of permanent continuous count stations, ATRs or similar, that log volume every day of the year. Rather than treat every one of those stations as its own universe, the traffic data unit sorts them into groups that share a similar month-to-month travel pattern: recreational and lake-access routes that spike in summer, commuter corridors that sag in August when school's out, ag and grain-belt roads with a harvest bump, urban arterials that barely move all year. Each group gets a set of monthly seasonal adjustment factors, usually built by averaging the ratio of monthly average daily traffic to annual average daily traffic across every station assigned to that group.

The factor table lives in the state's traffic monitoring guide or annual traffic volume report, sometimes as a GIS layer with a group ID attached to each route segment. The whole point is that you don't need a permanent counter on your link to know its seasonal shape. You just need to know which group it belongs to, because the group's pattern is assumed to be close enough to yours to correct for.

Running the annualizing math

A short-duration count, whether it's a 48-hour tube, a one-day pneumatic, or a single VHR pass, gives you a volume for one slice of the year. FHWA's framework for this is simple in principle: AADT equals your observed count divided by the applicable day-of-week factor, divided again by the seasonal adjustment factor for the month you counted and the group that link belongs to. Skip that seasonal step and you're implicitly claiming the day you counted represents the average day of the entire year, which is almost never true outside a handful of flat-pattern groups.

The size of the error depends entirely on how far your link's actual seasonal swing is from flat, and on whether you grabbed the right group. A beach-road count taken in July, left unadjusted, will read high for the year. The same road counted in February will read low. Both are wrong in opposite directions until the seasonal factor pulls them back toward the annual average.

This is where engineers actually get burned. A link doesn't announce its seasonal group; someone has to assign it, and the assignment is a judgment call based on land use, route function, and proximity to known seasonal generators. A rural connector near a state park gets lumped in with a commuter corridor more often than it should, usually because nobody walked the land use before pulling a factor off the nearest table. If your link sits on the boundary between two groups, read the group descriptions in the state's TMG rather than trusting the GIS layer blind. Some states publish a map of station locations by group, worth a look before you accept the default.

None of this changes how the raw count gets collected in the first place. You still need a number from the field, from a link that probably has no counter on it and no budget for one. That gap is what Vehicle Counting is built to close: one scheduled image pass over the corridor, delivered as a count by direction you can then run through whatever seasonal factor your DOT assigns to that link.

Get the seasonal group wrong and the best count in the world still produces a bad AADT. Get the count itself without sending a crew to lay tubes, and the seasonal math is the easy part left to do at your desk.

Start a pilot