Selection of sampling points for the Barcelona case study

The municipality of Barcelona is one of the three partners in the SCOREwater project. Other partners are the city of Göteborg (SE) and Amersfoort (NL). The Barcelona case focuses on ‘sewage sociology’, zooming in on health and environmental behaviours. Data will be extracted from three different sampling points, of which we will not disclose the location, since that could influence the research findings.

Selection of neighborhoods

Selecting representative sampling points is easier said than done, not only because Barcelona counts 73 neighbourhoods! The selection is made based on two sets of criteria, substantive and technical criteria. Substantive criteria refers to the type of neighbourhoods used for sampling, the socioeconomic status, which percentage is residential area, etc. Technical criteria refer to the actual measuring, like: “Is it possible to find a representative sampling point in the sewer which collects sewage from a given neighborhood without interferences from other neighborhoods?", "is it possible to install a monitoring station in the selected point?" In this post we will look into some of the substantive criteria that influenced the selection of sampling points.

To give an insight in the reasoning behind the selection of the sampling points, below some thought steps the team took.

• First of all we must hold into account what the number of households (the sampling unit) of the chosen neighborhoods is. We looked for sites with a well-defined sewer network, serving a bounded population and without interference from other sources, which led to relatively small populations. For small populations survey effort is high, meaning that the number of interviews with households that should be conducted in relation to the target population is high (representativeness of the sample). Thus, we had to be creative in finding balanced solutions.
• When selecting neighborhoods, it is important to look at differences in the socioeconomic and the health status of the inhabitants of the neighborhood; socioeconomic status influences both behavior and health. We looked for instance at different antibiotic consumption rates, obesity levels and cancer rates. This data is retrieved from the Catalan Agency for Health Quality and Evaluation (AQuAS) (http://observatorisalut.gencat.cat/en/inici/ ).
• Next to health status and socioeconomic status of the inhabitants, there are other criteria. It is important to hold into account the amount of restaurants, the amount of tourist accommodation and the presence of healthcare centers when selecting neighborhoods. First, having a large amount of restaurants will largely influence the concentration of oils and greases in the sampling points. Second, a high number of tourists in a certain area also can manipulate the data. This is because the sensors measure data that is not from inhabitants but from tourists. Therefore, the data is less reliable and representative. Lastly, neighborhood with healthcare centers such as hospitals have a lower chance of providing representative data, because there are more ill people in the area and therefore more traces of medicines in the sewage. These data from different sources were analysed by IERMB (https://iermb.uab.cat/en/). In conclusion, within the SCOREwater project we prefer to compare three neighborhoods with a small amount of restaurants and hospitals and few tourists to get the best insights!