Published 2nd August 2019

By Laura MacDonald, Research Assistant in our Neighbourhoods and Communities programme.

Our neighbourhood environments change and evolve often; some changes are minor, while others involve major transformation. Change can take various forms; green space created or removed, existing housing or amenities demolished, new housing estates built, new motorways created, or existing transport infrastructure modified or extended. Change may affect neighbourhood residents’ physical or mental health, or health-related behaviours, to their benefit or to their detriment. To study how change in our neighbourhoods might affect our health we need robust information but data showing how our neighbourhoods are changing, at a fine geographic scale, for the whole of Scotland, did not exist - until now. This is why we created the atlas and an interactive mapping application.

Background

The UK has some of the best longitudinal data (that is, where repeated observations of the same subjects are collected at various points over time to study change), about people’s lives and their health in the world. This information has proved incredibly useful in understanding health, including how differences in the health of the most and least deprived have developed over time, and how changes in peoples’ individual circumstances can affect their chances of good health.

In order to understand the role of neighbourhood in protecting or harming health, we also need longitudinal data on environment which we could join to these data on individuals. Some environmental characteristics, such as air pollution, are quite well captured over time but there is a particular gap in data about the built and natural environment. The built environment refers to man-made surroundings that provide settings for human activity; scales ranging from buildings to parks. There are various changes in the built and natural environment that happen and are recorded locally. However, there is not much data of this kind available at a national level. Existing available data includes static maps (see figure 1) where change is not quantified, or research where change is shown in small areas only (see figure 2).

What did we do?

Figure 1. OML data overlaid by grid

 OML data grid

OS Open Map Local data (OML) were downloaded from EDINA Digimap for Scotland for 2016 and 2017. OML is a free, detailed, street-level data mapping product (see figure 1). A Geographic Information System (GIS) was used to create small comparable units, called grids, for the whole of Scotland containing 500m by 500m grid cells. We calculated change in features over time within each grid cell. For 2016 and for 2017, buildings, roads, and woodland, were spatially joined to grid cells. For each cell we calculated:

  1. number of buildings
  2. sum of length of road section
  3. sum of area of woodland
  4. building, road and woodland change between time points

We then calculated the proportion of cells with changes in buildings, roads or woodland for Scotland as a whole, and by Council Area.

What did we find?

There were a number of changes within grid cell areas across Scotland:

  • 11.6% of cells lost/gained buildings
  • 12.2% lost/gained roads
  • 20.2% lost/gained woodland

For the majority of cells, changes were small; most cells lost/gained <10 buildings, <1000m of road, or <1000m² of woodland. An example of extensive change in all three features can be seen in the Google Earth images in figure 2. Between 2016 and 2017 a new estate, identified through our atlas, was built in this area in South Lanarkshire. Our method provides a way to highlight this type of new development that, until now, was difficult to find unless enquiring locally. There appeared to be a great deal of change within one year, it has therefore been valuable for us to develop this robust methodology to measure and quantify it.

Figure 2. 2016-2017 change in buildings, roads and woodland (Google Earth, 2019)

Neighbourhood change

What were the challenges?

In the atlas we reported on what appeared to be physical change but we cannot see from the information provided whether changes are physically real on the ground, are data errors, or due to features being re-categorised without physically changing (e.g. some paths in public parks were re-categorised as roads between 2016 and 2017). Figure 2 shows change in one area; it’s beyond the scope of this project to validate all changes. However, we have created the first large scale and comparable small scale dataset that pinpoints change in the environment, and the extent of that change across Scotland.

What’s next?

We hope to include additional environmental features, and look at change over a longer period of time. We will then join this information on change to health and health-related behaviour data, asking questions about the extent to which they have been affected. We can use our data to understand how, and which, specific changes within peoples’ neighbourhoods may disadvantage or benefit their health and behaviours, e.g. mental health, morbidity rates, death rates, road traffic accidents and active commuting etc.

Are you interested in learning more about this research?

If you have any questions about this research please contact laura.macdonald@glasgow.ac.uk

Download the atlas in PDF format. 

 

First published: 2 August 2019

<< 2019