Welcome to the Water quality section

Click on a tab below to get started.
If you have any questions contact us.

--> Read me first <--
Why did we do this project? Includes a brief history of the project and current regulations.

Why is this important?
According to the polluter pays principle, costs incurred in rectifying environmental pollution should not be born by the general public but by those responsible.

Large public events can affect bacterial levels in lakewater
There is no public testing for short term changes
As population density changes so does use of the lake shore
As the number of people increases so does the number of bathers
Better management
The organizers of events may not be aware of the effects.
Special thanks
Thanks to Rachel for her patience and expertise
Results year one and two
The sampling was repeated two years consecutively at the same locations, during the same time frame
Project summary "Microbiological survey"
In partnership with the hackuarium and biodesign.cc
Survey 2016
No of samples: 72
No of locations: 3
First sample: 2016-06-21
Last sample: 2016-08-09
Survey 2017
No of samples: 69
No of locations: 3
First sample: 2017-06-12
Last sample: 2017-07-31
Location of samples:
Map of sampling sites
The same sites were surveyed in 2016 and 2017
Results by year
Aggregated results by year, location and colony-colors
Click on image above to see results for 2017
Click on image above to see results for 2016
Economic and environmental influences
What are some of the factors that can cause an increase of bacteria in lakewater?
Large Events
The Montreux Jazz Festival
Second largest jazz festival in the world
Most of the events happen within 100 meters of the lake shore

The Montreux Jazz festival draws between 200K and 250K people per year. The organizers install temporary toilets that hook up to the local water treatment facility. As opposed to using chemical toilets.

The samples for this study were taken before, during and after the jazz festival.

Click on image for lager scale
Hotel nights sold per month
The number of hotel nights sold increases by 50% in one month
Source: Swiss federal office of statistics

Montreux is a tourist town, with a population of ~25'000. When this project was conceived, the annual report by the CIPEL showed that, sewage overflows were under estimated. See the report (in french)

The hotel nights effectiveley double (almost triple the population)

Click on image for lager scale
72 hour rainfall totals
Coliforms tend to increase in lake water during and after rain events
Increases in E.coli suggest a bypass of the water treatment system

If changes in the bacterial fauna are due to rainfall then we should see an increase in colony counts as rainfall increases.

However, the change would only be relfected if the capacity of the water treatment facility was surpassed. This would produce a spike in the graph above a certain amount of rainfall

Click on image for lager scale
Notebooks and code samples
Work in progress, analysis of results is done in a notebook first. This is a collection of those notebooks.
Data wrangling and cat herding
Putting the data together
The project was spread out over two years
Methods used to record the data changed from one year to the next

This is the initital notebook used to communicate interim results between team members

Here we decide on analysis methods, how to deal with results that are "Too numerous to count" or "Below detecable limit"

Links to a repository on Github

Aggregating the results
Developing the arrays of results
How to visualize all the weekly results from one year, by location and colony color.
Changes in the culture medium from one year to the next complicate the task just a smidge.

Results arrays are common when the data is generated by a machine, it is a whole different story when you need to code the array from scratch and account for non-numerical results.

Links to a repository on Github

Creating useable output
Creating charts and data for app and publications
Incorporating independent variables (hotel nights, rainfall)
Ensuring that print and application output are the same(not including color scheme)

Here we generate output for our webapp. That means creating dictionaries of values that can be exploited by a variety of platforms. We chose JSON as the format.

Links to a repository on Github