At this stage, there are a series of estimates which have been derived directly from count data available on that carriageway link. However, this obviously will not be the case for every link. There are then two scenarios left:

• Where we can infer an estimate from those adjacent links (this section)
• Where there is no relevant count history on or nearby to derive an updated estimate (next section)

There are several different methodologies available to infer what the estimate should be from those surrounding the link, often dependent on whether it is on one or both sides. There is also consideration of estimate along the road corridor and where this link can fit into that pattern.

Flows along a road will vary for various reasons and by various amounts as follows:

• There is a near continuous incremental change from property entrances along the route
• More significant variations come from trip generators such as schools or shops. Similar variations can come from side streets bringing traffic to or removing traffic from the road corridor
• Significant variations or step changes in volume can come from major intersections and bypasses
• There are also general patterns of increase and decrease as roads head towards the CBD for example or away from it. Similarly for rural roads, flows typically decrease as they head away from more populated areas and major routes.

These patterns need to be understood if realistic assessments are to be made from adjacent links. For example the estimate from an adjacent section may not be directly relevant to the link on the other side of a major intersection creating a step change in flows. Therefore any method selected below does require a manual check for appropriateness, and a degree of manual assessment anyway.

These methods are best used on longer roads with a number of links. The more links that have count derived estimates, the easier it is to update those links in between. As this is a typically manual type process, a focus on those roads with a at least five links and more than 30% having count derived estimates yields the best return for effort. These tend to be those higher volume routes that have a reasonable count programme.

Care should be taken before assessing a road corridor to ascertain any significant step changes in flow that effectively split the road corridor up into smaller combinations of links. The assessment of adjacent links should not cross these splits.

Recommended options include:

• Updating the latest estimate for the link to the current year by using the growth rate from the traffic group. Then checking the estimate aligns with other link estimates along the road corridor derived from count data and the asset manager’s understanding of the variations in flow along the route. This is the simplest method.
  • Alternatively, the average growth along the road corridor could be used if there are enough sections to support this analysis.
  • A similar method, particularly useful where estimate updates are regular, is to retain the proportional relationship between links from the previous estimate update.
• Taking the estimate from the adjacent link if appropriate and using that as the estimate for the link in question or averaging if between two links with count derived estimates
• Interpolating estimates between links along a road corridor. More complex techniques could utilise distance or chainage for interpolating estimates.

As mentioned earlier, this inferring of estimate data from adjacent or nearby sections is valuable where it can add value. But it does need manual oversight due to the reasons for the difference in flows between links.