Minimum user input data
Start with the amount of walking or cycling and the population represented by those data. The guidance below explains suitable data sources, how HEAT converts your inputs, and how to define the population.
Using counts from a trail or cycle track? See Using count data for HEAT assessments for an illustrated example covering trip distance, multiple counters, seasonal adjustment and rerouting.
Active travel data sources
Data on walking and cycling can come from different sources and may vary in quality. The data entered in HEAT should represent the average level of walking or cycling over a longer period; HEAT can also assess hypothetical scenarios that do not rely on observed data.
Population survey data
Population surveys can provide representative estimates of average walking or cycling across a population. These figures usually include people who do not walk or cycle, so the population size entered in HEAT should represent the whole population covered by the survey.
Large, regular surveys, such as national household surveys, are particularly useful. However, suitable data are not available everywhere since conducting a representative travel survey can be demanding. See also Population data inputs.
Short-term counts or surveys
Short-term counts can be strongly affected by weather, season, day of the week and time of day. A count on a sunny day, for example, may be very different from one on a rainy day.
Because HEAT requires long-term average levels, short-term counts may need to be adjusted. This is particularly important when assessing a specific facility, such as a cycle path or pedestrian bridge. An example for how this can be done is provided by the US National Bicycle and Pedestrian Documentation Project.
Continuous automated counts and large surveys conducted throughout the year are less affected by these issues.
Data from a few locations
Counts or surveys at only one or a few locations may not represent walking or cycling across a wider area. Results can vary depending on the location, season and type of users.
Such data should therefore generally be interpreted as representing the locations where they were collected rather than being automatically extrapolated to a whole city or community. HEAT allows adjustments to account for some spatial and seasonal differences.
Trip and count data
If you use the number of trips or people counted, you also need an estimate of the average trip length to calculate the total amount of walking or cycling. For example, a count on a bridge does not show how far people travel beyond the bridge.
Trip length can be estimated from surveys of users or from travel surveys.
For assessments based on counts, the population size does not affect the calculated result because the total amount of walking or cycling is already captured by the counts and trip length. You can nevertheless enter a population size that is relevant to the study area for reporting purposes; see Population data inputs.
HEAT currently does not support count data for motorized modes.
Bikeshare data
HEAT allows data from bikeshare systems to be entered in the flexible and full user interfaces.
Unlike most HEAT inputs, bikeshare data describe the whole system, rather than average use per person. Enter the average number of trips per day and average trip duration. Both are typically available from the bikeshare system.
Pedometer data
If you use step counts, make sure they mainly represent intentional walking, rather than all steps taken during the day.
Where possible, use data that distinguish walking from other steps. Alternatively, you can estimate the proportion of steps that represent intentional walking. A rate of around 100 steps per minute can be used to identify brisk walking.
How HEAT uses active travel data inputs
You can enter walking and cycling data in different ways, depending on the source. For example, you may have data on time, distance, trips, steps, frequency or mode share.
HEAT converts these inputs into common units so that they can be used in the calculations, mainly minutes or kilometres per person per day. Where needed, HEAT uses default values for conversions, such as average speed or trip distance. You can change these values where you have better local data.
How does HEAT convert different types of data?
- Time and distance: HEAT uses an average speed for each travel mode to convert between duration and distance.
- Steps: HEAT uses an average step length to estimate distance.
- Trips or counts: HEAT uses an average trip distance for the relevant travel mode to estimate distance.
- Mode share: HEAT combines the percentage of trips made by walking or cycling with the average total travel per person per day in the study area. It then converts the result into the required unit.
- Frequency: You can report how often people walk or cycle using five categories:
- daily or almost daily;
- 1–3 days per week;
- 1–3 days per month;
- less than once per month; or
- never.
For frequency data, HEAT estimates the number of days per year using the midpoint of each category. It then combines this with an estimated daily distance to calculate the average amount of walking or cycling per day.
Percent change in a two-case assessment
For two-case assessments, you can enter the comparison case as a percentage change from the reference case. For example, a 20% increase means that the amount of walking or cycling in the comparison case is 20% higher than in the reference case.
Population data inputs
HEAT assesses health impacts at the population level. Population size is therefore an important input, alongside the amount of walking or cycling. In general, the larger the population, the larger the estimated health impact.
The population data entered in HEAT must match the people represented by your walking or cycling data. This includes the population type, age range and population size.
Which population type should I use?
HEAT distinguishes between:
- General population: includes people who do and do not walk or cycle;
- Pedestrians: people represented by walking data; and
- Cyclists: people represented by cycling data.
The appropriate choice depends mainly on the source of your walking or cycling data.
Population surveys usually represent the general population, including people who do not walk or cycle. However, some surveys report average walking or cycling only among people who do some walking or cycling. In this case, the population size must be adjusted accordingly.
Count data include only people who walk or cycle and should therefore be treated as pedestrian or cyclist data.
The population type does not change the calculation itself. It mainly determines which input options are shown and helps ensure that you enter the appropriate population size.
Which age range should I use?
Age is important because mortality risk increases with age. The same amount of walking or cycling therefore generally results in a larger estimated health benefit in an older population.
Select the age range that best represents the people actually engaging in walking or cycling, even when you are assessing a general population.
| Walking | Cycling | E-biking* | |
|---|---|---|---|
| Young adult population | 20-44 | 20-44 | 20-44 |
| Adult population | 20-74 | 20-64 | 20-64 or 20-74 |
| Older adult population | 45-74 | 45-64 | 45-64 or 45-74 |
* Older age ranges are also available for e-biking because e-bikes are used relatively often by older adults.
HEAT excludes people younger than 20 because mortality is very low in this age group and there is insufficient evidence on the effects of walking and cycling on mortality (see also FAQ). It also excludes people older than 74 for walking and e-biking, and older than 64 for cycling. Mortality risks increase strongly at older ages, and including these age groups could substantially increase the estimated health benefits.
How do I determine the population size?
The population size entered in HEAT must match both the population type and the age range selected.
For country- and city-level assessments, HEAT provides a total population estimate where available, based on data from the United Nations Data Portal (1). It then estimates the proportion of the population within the selected age range using national data. You can review and overwrite these values if you have better local data.
If no suitable data are available, you need to enter the population size yourself before continuing.
For sub-city assessments, HEAT does not provide a default population size, so you need to enter the relevant population yourself.
Special case: count data
For assessments based on counts, population size does not affect the calculated health impact. The total amount of walking or cycling is already captured by the number of people counted and their estimated trip distance.
You can nevertheless enter a relevant population size for reporting purposes—for example, the population living near a pedestrian or cycling facility or within the area covered by the counts.
Sources:
1. United Nations. Data API. No year (https://population.un.org/dataportal/about/dataapi), accessed 27 October 2021. United Nations: New York.